How to Check if an Indian Judgment Is Still Good Law

A judgment that decided a matter when it was reported may be worth nothing today. Checking whether an Indian judgment is still good law is a distinct research step from finding it, and it is the step most easily compressed under filing pressure. This guide sets out what "good law" means in the Indian system, the four treatments to look for, the constitutional basis of binding precedent, and the separate question that damages many arguments: whether the statute the judgment construed still reads the way it did.

What "good law" actually means

"Good law" is not a status conferred by any court. It is shorthand for a working conclusion: as at today, this decision still states a proposition that a court is likely to apply, and it comes from a court whose word binds or persuades the forum you are appearing before. Three things can break that conclusion. Authority can fail, because a larger or higher bench has taken a different view. Scope can shrink, because later courts have confined the proposition to its own facts. Foundation can disappear, because the provision interpreted has been amended, repealed or legislatively reversed.

A crucial refinement: treatment attaches to propositions, not to case names. A judgment can remain excellent authority for its holding on limitation and be entirely displaced on its holding about jurisdiction. "Is this case good law?" is therefore the wrong question. The right question is "is this case still good law for the specific proposition I want to cite it for, before this forum?"

Article 141 and why the deciding court is the first check

Article 141 of the Constitution of India provides that the law declared by the Supreme Court shall be binding on all courts within the territory of India. Two phrases in that sentence do most of the work. "Law declared" is the operative expression, and its precise reach is discussed below. "All courts" includes the High Courts. The Indian hierarchy of authority runs as follows.

• The ratio of a Supreme Court decision binds every court and tribunal in India. The Supreme Court is not itself bound by its earlier decisions, but departs from them through a larger bench and with reasons.
• A High Court decision binds all courts and tribunals subordinate to it within its territorial jurisdiction. Before a different High Court it is persuasive only.
• Within a High Court, a Full Bench binds a Division Bench and a Division Bench binds a single judge. A bench of equal strength that disagrees must refer the question, not decide against the earlier view.
• Before a tribunal such as the ITAT, NCLT or CESTAT, a decision of the jurisdictional High Court binds. A decision of another High Court is persuasive where the jurisdictional High Court has not spoken.
• Strictly, it is the ratio that binds. The status of the Supreme Court's obiter is less settled: Indian courts have taken differing views on whether considered obiter of the Supreme Court is itself "law declared" under Article 141, and it is at minimum highly persuasive and routinely followed. Treat it as strong, but identify in your note whether the passage you rely on is ratio or obiter.

So the first check is not whether the case survives, but whether it would ever have bound your forum. Citing a persuasive High Court authority as though it were binding is a defect of the same order as citing an overruled case.

The four treatments and what each one costs you

Followed

A later court applies the proposition to comparable facts. This strengthens the authority, but the strength depends on who followed it. A Supreme Court bench following an earlier Supreme Court decision is worth far more than a single judge following it in a jurisdiction other than yours. Repeated following across benches and years is what turns a proposition into settled law.

Distinguished

The later court accepts the earlier authority but holds that the facts or the statutory context differ, so it does not apply. Distinguishing does not destroy the authority; it narrows it. What matters is the reason given for distinguishing. If the distinguishing feature is present in your matter, your authority will not travel. A case that has been distinguished repeatedly and followed rarely is usually a case that survives only on its own facts.

Doubted

A later bench records reservations about the correctness of the earlier view, or refers the question to a larger bench without deciding it. The proposition remains in force until it is displaced, but a pending reference is live risk. If you are relying on a proposition that is under reference, you should know that before opposing counsel tells the court.

Overruled

A competent bench holds the earlier proposition to be wrong. Overruling is often partial: a judgment may be overruled on one point and remain authoritative on the rest. Note the temporal effect. In the Indian system an overruling ordinarily operates retrospectively, the law being treated as always having been what the later court says it is, unless the court expressly limits the ruling to future operation. Prospective overruling has been available as a technique since I.C. Golaknath v. State of Punjab (1967), but it is exercised sparingly and must be declared.

Larger benches, coordinate benches and pending references

A bench cannot overrule a decision of a bench of equal or greater strength. Where it disagrees, the correct course is a reference to a larger bench. This produces a research hazard: a proposition can sit under a cloud for years while a reference is pending, technically binding but practically unsafe to build a case around without addressing the doubt.

It also means the status of a proposition can change quickly. In 2023 a five-judge Constitution Bench in N.N. Global Mercantile v. Indo Unique Flame took one view of the consequences of non-stamping for an arbitration agreement. Within months, a seven-judge bench in In Re: Interplay Between Arbitration Agreements under the Arbitration and Conciliation Act 1996 and the Indian Stamp Act 1899 took a different view, holding that non-stamping affects admissibility and is curable rather than rendering the agreement void. A Constitution Bench decision from earlier the same year was no longer good law by the end of it.

Per incuriam and sub silentio

A decision is per incuriam where it was rendered in ignorance of a binding statutory provision or a binding precedent, such that consideration of it would have produced a different result. A judgment held per incuriam by a competent court loses its precedential value. The plea is accepted sparingly, and a subordinate court cannot casually declare a superior court's ruling per incuriam in order to avoid it. Treat it as an argument to be made where the ingredients genuinely exist, not as a routine escape.

Sub silentio is narrower still. Where a point was assumed by the parties and the court rather than argued and decided, the decision is not authority for that point. Judgments can proceed on unexamined assumptions about jurisdiction, maintainability or limitation. Those assumptions are not precedent.

A citation is only as strong as the proposition it is offered for. Treatment attaches to propositions, not to case names.

The separate question: has the statute changed?

A frequent reason an Indian judgment stops being useful has nothing to do with later judgments. The provision it interpreted was changed. Four patterns recur.

• Amendment. The wording is altered, and an interpretation built on the earlier wording no longer holds. Always compare the text the court was construing with the text in force on your relevant date.
• Repeal and re-enactment. The Companies Act 2013 replaced the Companies Act 1956. The Bharatiya Nyaya Sanhita 2023, the Bharatiya Nagarik Suraksha Sanhita 2023 and the Bharatiya Sakshya Adhiniyam 2023 replaced the Indian Penal Code 1860, the Code of Criminal Procedure 1973 and the Indian Evidence Act 1872 with effect from 1 July 2024. The Income-tax Act 2025 replaced the Income-tax Act 1961. Older judgments remain valuable for interpretive principles, but section numbering and, in places, substance have moved. Do not assume a section number survives a re-enactment; locate the corresponding provision in the bare Act rather than guessing at it.
• Legislative reversal. Parliament may amend to neutralise a judicial outcome. Following Vodafone International Holdings B.V. v. Union of India (2012), the Finance Act 2012 amended the Income-tax Act 1961, then in force, with retrospective effect; that retrospective effect was later withdrawn by the Taxation Laws (Amendment) Act 2021. Because the 1961 Act has since been replaced by the Income-tax Act 2025, section numbers taken from the older Act should not be cited as current; identify and verify the corresponding provision in the governing income-tax legislation. Similarly, after Nikesh Tarachand Shah v. Union of India (decided 2017, reported (2018) 11 SCC 1) struck down the twin conditions in Section 45 of the Prevention of Money Laundering Act 2002, the provision was amended and the amended scheme was upheld in Vijay Madanlal Choudhary v. Union of India (2022).
• Subordinate legislation. Tax and regulatory rulings frequently turn on a notification, circular or rule since superseded. The judgment may be intact while the instrument it construed is not. Verify against the current official notification rather than a secondary summary.

This produces a discipline worth building into every research note: state the date as at which the law is applied, and record the version of the provision relied on. A judgment can be flawless on its reasoning and useless because the section it construed no longer exists in that form.

A summary SLP dismissal is not Supreme Court affirmation

A special leave petition under Article 136 of the Constitution invokes discretion. When the Supreme Court dismisses an SLP in one line, whether "dismissed" or "we see no reason to interfere", it is declining to exercise that discretion. It is not declaring law under Article 141, and it is not an endorsement of the High Court's reasoning. Treating it as one is an avoidable error in case notes and written submissions.

Kunhayammed v. State of Kerala (2000) sets out the distinction between the stages. Where leave is refused, the doctrine of merger does not apply and the order under challenge stands on its own footing as a decision of the court that passed it. Where leave is granted and the appeal is then decided, the lower court's order merges into the Supreme Court's judgment. And where an order, even at the SLP stage, contains reasons that state the law, those reasons carry the weight of a Supreme Court pronouncement to that extent. The practical rule: never rely on a bare "SLP dismissed" note. Read the order itself and see whether it says anything.

Why citation graphs make this tractable

Done manually, verification means pulling every subsequent decision that cites your case, reading each one, locating the passage where it is discussed, and classifying the treatment. That is feasible for a single authority and unmanageable for a brief resting on forty. It is also where the work silently degrades: the first three cases get checked properly and the rest get skimmed.

A citation graph converts the corpus into a network. Each judgment is a node and each citation is an edge. Where those edges are annotated with how the later court treated the proposition, several questions become visual rather than laborious.

• Whether the case has been followed by higher and larger benches or only by coordinate and subordinate ones.
• Transitive risk: your authority is untouched, but the decision it rests on has been overruled, which hollows it out.
• Clusters of distinguishing that show a proposition surviving only on narrow facts.
• Whether the question has been referred to a larger bench and is pending.
• Which of several competing authorities actually binds your forum, once results are filtered by court and jurisdiction.

LexVio's Indian court research covers the Supreme Court, High Courts, NCLT, ITAT, CCI and CESTAT, with citation graph visualisation and filtering by jurisdiction, so the citing relationships and the deciding forum are visible alongside the authority rather than assembled in a separate exercise. Reading the citing decisions and deciding how each one treats the proposition remains work for the professional. The graph narrows what a person has to read. It does not decide whether the case helps your client, whether to lead with it or pre-emptively distinguish it, or what to disclose to the court. That assessment, and the accountability that goes with it, stays with the professional.

A working checklist

• Identify the precise proposition you want the case for, not the case as a whole.
• Confirm the deciding court and whether it binds your forum.
• Confirm that the passage relied on is ratio, not obiter, and not a submission recorded by the court.
• Pull the subsequent citing history and classify treatments: followed, distinguished, doubted, overruled.
• Check for a pending reference to a larger bench on the same question.
• Compare the statutory text the court construed against the text in force on your relevant date.
• Check for amendment, repeal, re-enactment or a validating Act, and for superseded notifications and circulars.
• Where the trail includes an SLP, read the order rather than relying on a "dismissed" note.

Verification is not a formality performed after the argument is written. It is what determines whether the argument exists. Run it before you build on an authority, not after opposing counsel does it for you.