The Structure of the Indian Judiciary, Explained Properly

Almost everybody who has been through an Indian classroom can draw the judiciary from memory. A triangle. Supreme Court at the top, High Courts in the middle, district courts at the base. It is tidy, it is memorable, and it is close enough to the Constitution's text that nobody thinks to question it. It is also a poor description of how adjudication in this country actually happens, and if you use it as a map for finding law you will spend a lot of time looking in places where the answer is not.

The pyramid misleads in two directions at once. It understates how much internal variety sits inside the word "district courts", which is not one court but a graded family of civil and criminal forums with distinct powers. And it omits, entirely, the parallel structure of tribunals and special courts that now decides large parts of Indian commercial, tax, service and environmental law, and that does not sit under the High Courts in the neat way the triangle implies.

This article redraws the map. Not as constitutional theory, but as a working description of which forum decides what, what sits above it, and where the record of its decisions ends up.

What the Constitution Actually Sets Up

Start with the text, because the text explains the shape. Article 124 establishes the Supreme Court. Article 214 provides that there shall be a High Court for each State, and Article 231 permits Parliament to establish a common High Court for two or more States, which is why the number of High Courts has never matched the number of States. Articles 233 to 237 deal with the subordinate judiciary, and they do something the pyramid diagram never conveys: they place appointment, posting and promotion of district judges with the Governor in consultation with the High Court, and control over the subordinate judiciary with the High Court itself.

That last point is the load-bearing one. The Indian judiciary is a single integrated system, not a federal one. There is no separate stream of State courts applying State law and Union courts applying Union law. One hierarchy of courts applies the whole body of law, whichever legislature made it. An appeal from a magistrate in a small town can, in principle, travel all the way to a Constitution Bench without changing systems.

The Constitution also anticipated the parallel structure, though it is rarely read that way. Articles 323A and 323B, inserted by the Forty-second Amendment, expressly contemplate tribunals for service matters and for a list of specified subjects including taxation, industrial and labour disputes, land reform and elections. Tribunalisation is not a deviation from the constitutional design. It is part of it.

The District Judiciary Is Not One Court

The base of the pyramid is drawn as a single block labelled district courts. In reality each district contains two graded ladders, civil and criminal, that run in parallel and are usually staffed by the same cadre of officers wearing different hats.

The civil ladder

At the bottom sits the Civil Judge, Junior Division, competent to try suits up to a pecuniary limit fixed by State legislation. Above that, the Civil Judge, Senior Division, with a higher or unlimited pecuniary competence depending on the State. Above that, the District Judge, who is both the highest civil court in the district and the appellate forum for decrees passed by the courts below. Nomenclature varies: Munsiff in some States, Subordinate Judge in others, Additional District Judge for the many officers of District Judge rank who are not the principal district judge.

The practical point is that pecuniary jurisdiction is State law, not central law, and it is revised periodically. A suit that must be filed before a Civil Judge Senior Division in one State may belong to a Junior Division court across the border. Anybody researching a procedural point about the correct forum has to read the relevant State's civil courts legislation, not a general proposition.

The criminal ladder

Criminal courts run from Judicial Magistrates, through Chief Judicial Magistrates, to Sessions Judges and Additional Sessions Judges, with the sentencing powers of each fixed by the procedural code. The Bharatiya Nagarik Suraksha Sanhita, 2023, which replaced the Code of Criminal Procedure, 1973, retained this ladder but simplified it in one visible respect: it did away with the separate classification of metropolitan areas and metropolitan magistrates, so the older vocabulary you will still meet in judgments and case titles describes a category that no longer exists in the current statute.

Sessions courts also carry a large volume of special jurisdiction. Courts designated as Special Courts under the Protection of Children from Sexual Offences Act, 2012, under the Prevention of Money Laundering Act, 2002, under the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989, and under several other statutes are very often courts of session on which an additional designation has been conferred. That fact matters enormously for both workload and data, and it gets a section of its own below.

High Courts: Constitutional Courts With Two Different Jobs

The High Court is where the pyramid is least misleading and still slightly wrong. There are twenty-five High Courts, and their territorial reach follows history rather than a rule. Some serve one State, some serve a State and a Union Territory, some serve several States. Several have permanent benches at cities other than the principal seat, with the distribution of work between seat and bench governed by notifications and by orders of the Chief Justice rather than by anything a researcher can infer from the map.

What matters for research is that a High Court does two structurally different things. It sits at the top of the ordinary appellate chain, hearing first appeals, second appeals and criminal appeals from the district judiciary. Separately, and independently of that chain, it exercises writ jurisdiction under Article 226 and superintendence over all courts and tribunals within its territory under Article 227. The second job is what pulls tribunals back under the High Court even when the tribunal's own statute routes appeals elsewhere.

A handful of High Courts additionally retain ordinary original civil jurisdiction, a legacy of the chartered High Courts, so that suits above a pecuniary threshold are filed in the High Court itself rather than in a district court. The Delhi High Court's threshold was raised to two crore rupees by amending legislation in 2015. Anyone researching commercial litigation in those cities needs to know that a first-instance suit and a second appeal can both be High Court matters, decided by the same institution in completely different capacities.

A High Court is not one court doing one job. It is an appellate court, a writ court, a supervisor of every tribunal in the State, and in a few cities a trial court as well.

Where Tribunals Actually Sit

This is the part the pyramid gets flatly wrong. Tribunals are usually drawn, if they are drawn at all, as a small box hanging off the side. In practice they carry entire fields of law from first instance to final fact-finding, and their relationship to the ordinary hierarchy differs from statute to statute.

There are broadly three patterns, and knowing which one you are in tells you where to look for both the governing authority and the record.

Whichever pattern applies, one principle sits above all of them. In L. Chandra Kumar v Union of India, decided in 1997, the Supreme Court held that the power of judicial review vested in the High Courts under Article 226 and 227 is part of the basic structure of the Constitution, and that tribunal decisions remain subject to scrutiny before a division bench of the High Court within whose jurisdiction the tribunal falls. Tribunals can substitute for courts as forums of first instance. They cannot be walled off from the constitutional courts.

Special Courts, Designations, and the Nameplate Problem

Alongside tribunals sits a third category that confuses the picture further: courts that are ordinary courts wearing a statutory designation. A commercial court under the Commercial Courts Act, 2015, a special court under POCSO, a designated court under the National Investigation Agency Act, 2008: in most cases these are not new institutions with new judges and new buildings. They are existing district judiciary courts on which a statute has conferred exclusive jurisdiction over a category of case, with a procedure the statute prescribes.

This has two consequences worth naming. The first is a capacity illusion. Creating a special court by designation adds jurisdiction without adding a judge, so the same officer now carries the special docket and whatever else was already listed. The second is a data consequence. The same physical court appears in records under multiple identities depending on the hat it wore that day, and case type codes for special jurisdictions vary between States, which makes counting anything across the country harder than it should be.

Then there are genuinely separate institutions created by statute at the district level. Each has its own composition, its own procedure and its own publication practice, and none of them fits inside a triangle.

• Family Courts under the Family Courts Act, 1984, which take over matrimonial, maintenance and custody jurisdiction in the areas where they are established.
• Juvenile Justice Boards under the Juvenile Justice (Care and Protection of Children) Act, 2015, constituted with a magistrate and two social workers rather than a judge alone.
• Consumer commissions at district, State and national level, with their own pecuniary tiers and their own appellate chain.
• Permanent Lok Adalats for public utility services, which uniquely can decide on merits where conciliation fails, subject to a pecuniary limit.
• Motor Accident Claims Tribunals, which handle a very large civil docket under the motor vehicles legislation and are usually presided over by district judiciary officers.
• Debt Recovery Tribunals, which sit outside the district judiciary entirely and answer to their own appellate tribunal.

The Parallel System Nobody Draws At All

One more structure deserves mention because it is genuinely invisible in most descriptions: the revenue courts. Across most States, a hierarchy running from the tahsildar through the sub-divisional officer and the collector to a divisional commissioner and a board of revenue or financial commissioner decides mutation, partition of agricultural holdings, tenancy questions and land record corrections under State land revenue legislation.

These are not part of the judiciary in the constitutional sense. They are executive officers exercising quasi-judicial powers, and their decisions are frequently subject to a statutory bar on civil court jurisdiction in the matters they cover. For the most common category of dispute in India, which is a dispute about land, the first forum is very often not a court at all. The record of what these forums decide is the thinnest and least standardised legal data in the country.

Why This Map Is a Research Skill, Not Trivia

The reason to hold an accurate map is that every research question begins with an implicit forum assumption, and a wrong assumption produces a confident, empty answer.

What a Unified View of the Structure Changes

Once you accept that Indian adjudication is a hierarchy plus a parallel tribunal system plus a set of statutory designations plus a revenue structure, the research problem changes shape. It stops being a question of finding the best database for judgments and becomes a question of covering the right forums at once.

That is the reasoning behind how CourtMesh was built. One search runs across the Supreme Court, all twenty-five High Courts, the district judiciary and tribunals including the NCLT, NCLAT, ITAT and CESTAT, with roughly 310 million records keyword-searchable and roughly 2 million carrying deeper semantic indexing. Everything is drawn from official government portals, so what you are reading is the published record rather than an intermediary's version of it.

It does not dissolve the structural problem, and we would rather say so than pretend. Publication practice differs by forum, revenue courts remain largely outside any consolidated index, and a nil result never establishes that a matter does not exist. What a single window changes is the cost of asking a question across the whole structure rather than in one corner of it, which is exactly the cost that a wrong mental map imposes on you without ever telling you it has.