Summary. A state supreme court grants review in a small fraction of the cases presented to it, and the reasons it grants have little to do with whether the decision below was wrong. Discretionary review exists to resolve conflicts, settle questions of first impression, and correct decisions of institutional significance, which means a petition arguing that the intermediate court misapplied settled law to particular facts is asking for something the court does not do. This guide covers what the criteria actually require, how to preserve and frame an issue so that it is reviewable at all, the mechanics and deadlines that vary by state, the drafting of the petition and the answer, and what happens after review is granted. It also addresses the strategic questions — federal claims, amicus support, publication practice, and stays — that shape the outcome before any brief is written.

The intermediate appellate court has affirmed. The client is unhappy, believes the panel got it wrong, and wants to know what comes next.

What comes next is a petition that will almost certainly be denied. State supreme courts grant review in somewhere between three and fifteen percent of civil petitions depending on the state, and the denial rate is not a measure of how many petitions are meritless. It is a measure of the court's function.

That function is the thing to understand before writing anything. A state supreme court is not a second court of appeals. It exists to make law for the state — to resolve conflicts among intermediate districts, to answer questions nobody has answered, and to correct decisions whose consequences reach beyond the parties. A petition that says the panel misapplied settled law to these facts is describing precisely the case the court declines to take, however right it may be.

What the criteria actually mean

Most states enumerate criteria closely tracking U.S. Supreme Court Rule 10, and most enumerate them as considerations rather than requirements.

Conflict

The strongest ground. A decision that conflicts with a decision of another intermediate district, with a prior decision of the supreme court, or with a decision of another appellate panel presents exactly the problem the court exists to solve — the same law meaning different things in different parts of the state.

To make it work:

  • Quote both decisions. Not characterizations — the actual language, side by side, on the same legal question.
  • Show the conflict is real, not distinguishable on the facts. A court will not take a case to resolve a conflict that does not exist.
  • Show it is outcome-determinative here.
  • Show it recurs. A conflict that arises once every decade is less compelling than one that district courts face weekly.

In several states, an intermediate court's certification that its decision conflicts with another is itself a basis for review, and in a few it creates an appeal as of right. Ask the panel to certify.

Question of first impression

A legal question the state's courts have not decided. Strong where the question recurs, where lower courts are visibly uncertain, and where the answer matters to people other than these parties.

Weaker than practitioners assume when the question is genuinely novel but also genuinely rare. Courts prefer to decide questions that are ripe, well-developed below, and presented on a clean record.

Public importance

The decision affects a substantial class of persons, a significant public interest, or the operation of government. This is where amicus support does its real work — a brief from a trade association, a municipal league, or a bar section demonstrates that the question matters beyond the caption.

Departure from accepted judicial practice, or a decision so erroneous as to require correction

Most states include a residual ground for error correction. It is available and it is rarely the reason review is granted. Where the error is genuinely stark — a decision that misreads a statute's plain text, or that overlooks a controlling precedent — lead with it; where the argument is that the panel weighed the evidence wrongly, do not file.

Constitutional questions

A decision invalidating a statute, or resolving a significant state or federal constitutional question, draws attention in most courts and in several states carries a right of review.

Before the petition: preserving something to review

Review is not available for an issue that was never presented. Most of the work that determines whether a petition can succeed happens years earlier.

Preserve in the trial court. An objection, a motion, an offer of proof, a proposed instruction with an objection to its refusal. The specific ground must be stated; a general objection preserves nothing.

Raise it in the intermediate court, expressly. An issue not briefed there is generally waived, and most states will not entertain an issue raised for the first time on review. Where an issue was raised but not addressed by the panel, say so and point to the briefing.

Get the record right. Discretionary review is decided on the record made below. A factual assertion unsupported by the record is worse than useless.

Consider whether to seek rehearing. In many states a petition for rehearing in the intermediate court is a prerequisite to review, or is necessary to preserve an issue the panel decided without briefing. In others it merely delays. Check the rule, and note that rehearing frequently extends the deadline to petition — which is sometimes the only reason to file one.

Watch for the federal question. If a federal issue may eventually go to the U.S. Supreme Court, it must be presented to the state's highest court first under 28 U.S.C. § 1257. And the state court's decision must not rest on an adequate and independent state ground, which would defeat federal review entirely under Michigan v. Long, 463 U.S. 1032 (1983). Where a state constitutional claim is pleaded alongside a federal one, understand that winning on the state ground may preclude federal review — sometimes desirable, sometimes not.

Mechanics and deadlines

These vary by state more than any other aspect, and the rule must be read rather than remembered.

Nomenclature. Petition for review, petition for allowance of appeal, application for leave to appeal, petition for writ of certiorari, or notice of appeal where review is of right. The name signals nothing about the standard.

Timing. Commonly ten to thirty days after the intermediate court's decision or after denial of rehearing, and in several states the periods are jurisdictional and cannot be extended. Some states permit an extension on motion filed before expiration; some do not.

Length. Strict page or word limits, frequently much shorter than a merits brief — ten to twenty pages is typical. This is a feature: the court is deciding whether the case is worth its time, and a petition that cannot make that case briefly probably cannot make it.

Format. Required sections vary but generally include the questions presented, a statement of the case, the decision below, the reasons review should be granted, and an appendix containing the opinions below.

Filing fee and service, and in many states service on the intermediate court and the trial court.

Answer. The respondent typically has a fixed period to answer, and in most states answering is optional. Whether to answer is a real strategic question, addressed below.

Reply, where permitted, usually short.

Amicus briefs at the petition stage, permitted in most states with the court's leave or the parties' consent, and disproportionately valuable.

Cross-petitions, where the respondent also lost something below.

Certified questions. Several states permit a federal court, or another state's court, to certify a determinative question of state law. This is a distinct and often superior route: a certified question arrives with no procedural baggage, is framed by a court rather than a party, and is accepted at much higher rates.

Direct review. Some states permit bypass of the intermediate court in cases of unusual public importance or urgency, on petition before the intermediate court decides.

Writing the petition

The questions presented

The most important part, and the part written last after the rest is drafted.

  • Short. One or two sentences each. A question presented that takes a paragraph has smuggled in an argument.
  • Framed as legal questions, not as complaints about the outcome. "Whether a residential landlord's duty of care extends to a common area maintained by a third-party contractor" — not "whether the court of appeals erred in affirming summary judgment."
  • Framed at the level of generality that makes the case important. Too narrow and it is fact-bound; too broad and it does not arise on this record.
  • Answerable on this record. If the court would have to assume a fact not found below, the question is not presented.
  • Few. Two questions, occasionally three. A petition raising seven questions tells the court that none of them is compelling.

The reasons for granting review

Lead with the criterion, not the merits. The first heading should be why the court should take the case — the conflict, the recurring question, the public consequence — and the merits should appear in service of that.

A petition organized as a merits brief with a certiorari section appended is the most common defect, and it reads as a request for a third opinion.

Demonstrate recurrence. Cite the other cases presenting the question. Cite the trial court decisions going both ways. Cite the treatise noting the split, the law review article, the practice guide that flags the uncertainty. Show that the court's answer will matter next month in other cases.

Show the consequences concretely. How many contracts are drafted against the rule below? How many claims will be barred? What will regulators or lower courts do without guidance?

Make the merits argument briefly and confidently. The court will not grant review to affirm, so the petition must make the reader think the decision below is probably wrong — but this section should be compressed, not comprehensive. Save the full argument for the merits brief.

Be scrupulous about the record and the authority. A petition that overstates a holding, or that cites a case for a proposition it does not support, loses on credibility, and appellate staffs check.

The statement of the case

Short, factual, and fair. Include the procedural history, the disposition below, and the facts that make the legal question concrete. Adverse facts should appear here rather than in the answer.

The appendix

The opinions below, the trial court's dispositive order, the operative statute or provision, and any conflicting decisions quoted in the petition.

Whether to answer

The respondent won below. The instinct is to explain why the petition is wrong.

In most states the answer is optional, and in the ordinary case declining to answer is reasonable — it signals that the case is unremarkable, and it avoids drawing attention to arguments the court might otherwise skim past.

Answer when:

  • The petition misstates the record or the holding below, which happens more than it should and which a court will not catch on its own.
  • There is a vehicle problem — waiver, an alternative ground for affirmance, mootness, an interlocutory posture, a factual dispute that would prevent the court from reaching the question. Vehicle problems defeat more petitions than merits arguments do, and they are invisible unless someone points them out.
  • The claimed conflict is not real, and the distinction can be shown in a paragraph.
  • The court requests an answer, which in most states is a strong signal of interest and requires the respondent's best work.
  • The case is important enough that a grant would be costly regardless of the ultimate outcome.

The answer should be shorter than the petition and should be organized around why review is unwarranted rather than why the respondent wins.

Consider a conditional cross-petition where the respondent has an issue it would want reached if review is granted.

Amicus support at the petition stage

Disproportionately valuable, and underused in state practice.

An amicus brief at the petition stage answers the question the court is actually asking — does this matter to anyone besides these parties — in a way the party cannot. A brief from a trade association, a bar section, a municipal league, a group of law professors, or an affected industry says: we face this question, the uncertainty costs us, and we need an answer.

Practical guidance:

  • Recruit early. Organizations move slowly and need lead time, board approval, and often a committee review.
  • Coordinate but do not duplicate. Amici should each contribute a distinct perspective; three briefs saying the same thing are worth less than one.
  • Focus on consequences, not merits. The party briefs the law. The amicus explains why the answer matters.
  • Check the rules on consent, leave, disclosure of authorship and funding, and timing. Some states are strict.
  • Do not manufacture it. A court recognizes an amicus brief written by the party's counsel and funded by the party, and it counts against the petition.

Publication and depublication

A quiet but consequential area in the states that have it.

Where an intermediate court issues an unpublished decision, it generally has no precedential effect and may be uncitable. A party harmed by an unpublished decision may move to publish it — which is sometimes desirable, because an unpublished decision cannot create a conflict and therefore cannot support review on that ground.

Conversely, several states permit depublication: the supreme court orders the intermediate decision unpublished without granting review, removing its precedential effect while leaving the judgment intact. For a party whose real concern is the rule rather than the outcome, a depublication request is a cheaper and more attainable alternative to review, and it should be considered as a fallback in every petition where the client's interest is prospective.

Where publication rules matter, address them expressly: ask for review, and in the alternative for depublication.

After a grant

Review is frequently limited. The order may grant as to one question and deny as to others, or may reframe the question. Read it precisely; briefing outside the granted issues is disregarded and irritates the court.

Full merits briefing follows on a new schedule, with new length limits. The petition is not the brief; the merits brief is written from scratch, at length, with the full argument the petition compressed.

Amicus participation expands at the merits stage and should be coordinated.

Oral argument is typically granted. State supreme courts sit en banc, argument is often longer than in intermediate courts, and questioning is directed at the rule the court will announce rather than at the record. Prepare for the question "what should the rule be, and how would it apply in the case you have not thought about."

Consider settlement. A granted case has value to both sides that it did not have the week before — and a party facing an adverse rule of statewide application may pay well to avoid it. Note that several courts will not dismiss a granted case on stipulation where the question is important, and may decide it anyway.

Practical judgment: when not to petition

Counsel does a client a service by saying no.

Do not petition where the argument is that the panel weighed the evidence wrongly, applied settled law to particular facts, or reached an unfair result on a discretionary ruling. These are the merits of an appeal that has already happened.

Do not petition where the issue was not preserved, where the record does not support the question, or where an alternative ground for affirmance means the court could not change the outcome.

Do not petition merely to delay, unless a stay is genuinely needed and the client understands the cost. Note that filing does not automatically stay enforcement; a supersedeas bond or a stay order is generally required, under the intermediate court's or the supreme court's rules.

Do consider the alternatives: a rehearing petition where the panel overlooked something specific, a depublication request where the concern is precedential, a certified question where the case is in federal court, and legislative advocacy where the answer the client wants is a change in the law rather than an interpretation of it.

And tell the client the odds. A petition costs fifteen to forty thousand dollars and is denied roughly nine times in ten. Clients who understand that before filing are not disappointed by a one-line order; clients who were told the case was strong are.

Primary authority

  • State rules of appellate procedure governing discretionary review — the operative authority, differing by state in nomenclature, deadlines, length limits, required contents, and whether rehearing below is a prerequisite.
  • U.S. Supreme Court Rule 10 — the criteria most state provisions track.
  • 28 U.S.C. § 1257 — U.S. Supreme Court review of final judgments of the highest state court in which a decision could be had.
  • Michigan v. Long, 463 U.S. 1032 (1983) — the adequate and independent state ground doctrine and the plain-statement rule.
  • Cox Broadcasting Corp. v. Cohn, 420 U.S. 469 (1975) — finality for purposes of § 1257.
  • State certification statutes and rules permitting federal courts to certify determinative questions of state law.
  • State publication and depublication rules, and rules governing citation of unpublished decisions.
  • Fed. R. App. P. 32.1 — citation of unpublished federal decisions, for the federal analogue.
  • State amicus rules on consent, leave, disclosure, and timing.

Two petitions, side by side

The one that is granted

A construction subcontractor's mechanic's lien is held invalid because the notice was served on the general contractor rather than the owner. The intermediate court's Third District holds that the statute's notice provision is strictly construed. Two years earlier the First District held on materially identical facts that substantial compliance suffices.

The petition writes itself.

Question presented: Whether the notice requirement of the mechanic's lien statute is satisfied by substantial compliance, or requires strict compliance.

Reasons for review, first heading: the districts are in direct conflict. Both opinions are quoted at length, on the same statutory provision, reaching opposite results on indistinguishable facts. Neither is distinguishable and the Third District expressly acknowledged the disagreement.

Second heading: the question recurs constantly. Every construction project in the state generates lien notices. Four trial court decisions are cited going both ways. The state's construction bar association files an amicus brief at the petition stage saying that its members cannot advise clients which rule applies and that title companies are refusing to insure over liens as a result.

Third heading, compressed to two pages: the Third District's reading is inconsistent with the statute's remedial purpose and with the legislature's 2019 amendment.

Review granted. Not because the panel was wrong, but because two districts cannot both be right and the answer matters to everyone building anything.

The one that is denied

A jury returns a defense verdict in a personal injury case. The intermediate court affirms, rejecting arguments that the trial court improperly excluded a treating physician's causation opinion under the state's expert rule and improperly instructed on comparative fault.

The petition presents four questions, each beginning "whether the court of appeals erred." The reasons section argues the evidentiary ruling was an abuse of discretion, that the instruction misstated the standard, and that the cumulative effect denied a fair trial.

Every one of these may be correct. None is a reason for review. The expert ruling is discretionary and fact-bound. The instruction issue was resolved by applying settled law. There is no conflict identified, no question of first impression, and no consequence beyond the parties.

What counsel should have said at the outset: the appeal is over, the odds of review are under five percent, a petition will cost twenty-five thousand dollars, and the money is better spent elsewhere. If the client wants to proceed anyway — and some will, and are entitled to — the petition should be narrowed to the single strongest issue and should lead with whatever institutional argument exists, however thin, rather than with four assignments of error.

Stays and security while the petition is pending

Filing a petition does not stop the judgment. A prevailing party may begin executing the day the intermediate court's mandate issues, and a client who believes the case is still pending can find its accounts garnished.

The mandate is the event to watch. Most intermediate courts stay issuance of the mandate for a period after decision, and a timely petition for rehearing or for review often extends that stay automatically. Read the rule; in several states it is automatic, in others it requires a motion, and in a few nothing stays the mandate absent an order.

Supersedeas. A stay pending review ordinarily requires a bond or other security in the amount of the judgment plus interest and costs. Several states cap the bond amount by statute — a response to appeals rendered impossible by nine-figure verdicts — and some permit alternative security such as a letter of credit or a lien on identified assets.

Where the bond is unaffordable, options include moving to reduce or waive it on a showing of the party's financial condition and the absence of prejudice, offering substitute security, negotiating a standstill with the judgment creditor, or seeking a partial stay limited to specific assets. Courts have discretion, and a well-documented motion showing that execution would destroy the business is sometimes granted.

Non-money judgments. An injunction, a specific performance decree, or an order affecting custody or possession requires a discretionary stay under the traditional factors — likelihood of success, irreparable harm, balance of harms, and the public interest. The likelihood-of-success factor is harder here because two courts have already ruled, so the argument must rest on the strength of the ground for review rather than on the merits generally.

Move in the right court, in the right order. Most rules require applying first to the intermediate court and only then, on denial, to the supreme court. Skipping the first step results in denial without consideration.

Move early. A stay motion filed after execution has begun is asking the court to undo something, which is a materially harder request than preventing it.

Interest keeps running. A stay preserves the status quo; it does not suspend post-judgment interest. Where the rate is high and the prospect of review is thin, the cost of delay may exceed what a resolution would cost today — a calculation worth putting in front of the client in writing.

Certified questions: the better route when it is available

A federal court facing an unsettled question of state law that will determine the case may certify it to the state's highest court. Nearly every state permits this by statute or rule, and it is the most efficient path to a state supreme court decision that exists.

Why acceptance rates are so much higher. A certified question arrives without a petition's baggage. There is no vehicle problem, because the certifying court has already determined the question is outcome-determinative. There is no waiver problem, because the federal court frames the question. There is no record dispute, because the certifying order recites the facts as the federal court finds or assumes them. And the state court is being asked to do the thing it most likes doing — declare state law — without having to review anyone's judgment.

How to get one. Move the federal court to certify, or raise it in briefing. The standards typically require that the question be determinative of the case and that there be no controlling precedent in the state's decisions. Lehman Brothers v. Schein, 416 U.S. 386 (1974), encourages certification where state law is uncertain, and Arizonans for Official English v. Arizona, 520 U.S. 43 (1997), describes it as a means of avoiding both unnecessary constitutional adjudication and the friction of a federal court guessing wrongly.

Draft the question yourself. Federal courts routinely adopt a party's proposed formulation. A well-framed question — narrow enough to answer, broad enough to matter, and neutral in phrasing — is worth considerable effort. A question that presupposes the answer will be reformulated or declined.

Timing. Certification adds time — commonly six to eighteen months — and the federal case is stayed. That delay is a cost to a party seeking speed and a benefit to one that is not, and the strategic implication should be assessed honestly rather than dressed up as neutral procedure.

State-to-state certification. Several states permit certification from another state's highest court, which arises where a forum applies a sister state's law to a genuinely open question.

The parallel move in an ordinary state appeal is asking the intermediate court to certify a conflict with another district, which in several states creates a right of review or substantially improves the odds. Ask for it in the rehearing petition; it costs a paragraph.

Building the record for review from the beginning

The best petitions are made possible by decisions taken years earlier, usually by trial counsel who was not thinking about a supreme court.

Frame legal issues as legal issues. A motion in limine arguing that particular testimony is unfairly prejudicial produces a discretionary ruling reviewable for abuse of discretion — unreviewable in practice. The same objection framed as a question about what the governing rule is produces a legal ruling reviewed de novo. The difference is often available and is rarely taken.

Get a ruling. An objection the court reserves and never decides preserves nothing. Ask for a ruling, and if the court declines, say on the record that the objection stands and the request for a ruling was denied.

Make the offer of proof. Excluded evidence is unreviewable without a record of what it would have shown.

Submit proposed instructions in writing and object specifically to refusals. "We object to the refusal of our instruction 14" preserves; "we object" does not.

Use special verdicts or interrogatories where liability rests on alternative theories. A general verdict on multiple theories is affirmed if any theory is supportable, which insulates the error.

Create the conflict expressly. Where the controlling district's rule differs from another district's, say so in the trial brief and in the appellate brief, citing both decisions. A conflict identified below is far more persuasive than one discovered at the petition stage, and it invites the panel to address or certify it.

Watch the alternative grounds. An appellee will defend the judgment on any ground supported by the record. Counsel who preserves only one issue while three alternative grounds for affirmance sit unaddressed has built a vehicle problem into the case.

Keep the record clean and complete. Order the transcripts. Ensure exhibits are admitted and included. A designation that omits the hearing where the ruling was made is a recurring and fatal error.

Think about the rule you want announced. A supreme court decides what the law is. Counsel who can articulate, in one sentence, the rule the client wants and how it would apply in the cases that will follow is prepared for the only question the court is really asking — and that sentence should exist from the first motion, not from the petition.

Working with appellate counsel

Discretionary review is a specialty, and the decision whether to bring in someone who does it regularly is worth making deliberately rather than by default.

What separate appellate counsel adds. Distance from the case, which matters enormously — trial counsel who lived the facts for three years finds it genuinely difficult to see that the compelling injustice is not a certworthy question. Familiarity with the particular court's grant patterns, which are idiosyncratic and largely undocumented. And credibility with a court that recognizes the appellate bar.

What trial counsel knows that appellate counsel does not. The record, exhaustively. Where the preservation is, and where it is not. Which alternative grounds for affirmance are real. The best arrangement pairs them, with trial counsel supplying the record and the honest assessment of its weaknesses and appellate counsel framing the question.

The uncomfortable conversation. Sometimes the strongest available argument is that trial counsel erred — a waiver, a failure to object, an issue not raised. A firm that handled the trial has an obvious conflict in evaluating that, and clients are entitled to advice from someone who does not.

Timing. Bring appellate counsel in before the intermediate court briefing if the case is headed for review, not after the loss. Issues can be framed for certworthiness during the appeal; they cannot be reframed afterward.

Costs. A petition and answer runs fifteen to forty thousand dollars in most markets. Full merits briefing and argument after a grant runs substantially more. Where a trade association or industry group has an interest, cost-sharing arrangements and amicus support can defray it, and those conversations take months to arrange.

The client's decision, informed properly. Present the odds, the cost, the timeline, the exposure while the case remains open, and the alternatives — rehearing, depublication, legislative advocacy, or acceptance. A client who chooses to petition with that information has made a reasonable decision even when the order denying review arrives four months later on a single line.

Related articles

This guide is provided for general informational purposes and does not constitute legal advice. Discretionary review procedures differ substantially among the states, including deadlines that are frequently jurisdictional and cannot be extended, whether rehearing below is a prerequisite, and whether depublication or certification is available. Filing a petition does not stay enforcement of the judgment. Consult counsel admitted in the relevant state and read the current appellate rules before filing.