What We Found Verifying Land Documents in India: An Honest 2026 Casebook

How we counted (method and sample)

Each document set was read by an OCR and reasoning pipeline and run through five independent checks: title-chain continuity, encumbrance, owner consistency, plot identity, and registered-transaction history. Of 135 verifications, 59 had all five checks resolve to a definite outcome (the rest were partial). We report only pooled percentages over that set — no individual file, location, name, or parcel is identified or identifiable here. The sample is self-selected: people who verify are, by definition, people with a reason to. Read every figure below as "among documents brought in for checking," never "of all property in India."

Finding 1 — Only 8% had a clean title chain; 73% had a break or gap

India operates a presumptive, not conclusive, title system: registration under Section 17 of the Registration Act, 1908 records that a transaction happened — it does not guarantee the seller owns what they are selling. Ownership has to be traced document-to-document, backward in time. That chain breaks for ordinary, non-fraudulent reasons: the Sabik-to-Hal khata transition in Odisha breaks the link on a large share of older plots; West Bengal's CS to RS to LR record systems create three-way reconciliation gaps; and a single un-recorded mutation anywhere in the chain leaves a hole. Hypothetically: a buyer is shown a crisp current record-of-rights, but the parcel's prior-owner entry never carried forward from the old settlement record — the document looks clean, the chain is not.

Finding 2 — 63% had an undisclosed or unresolved encumbrance

A sale deed describes a transaction; it does not reveal whether the land is mortgaged, liened, or subject to a court attachment. That history lives in the Encumbrance Certificate (Form 25) — and a deed that reads perfectly can sit over an unreleased mortgage. Hypothetically: a plot was pledged for a loan years ago, the loan was repaid, but the release deed was never registered — so the charge still legally clouds the title until it is cleared. A buyer reading only the sale deed would never see it.

Finding 3 — 63% had an owner-name mismatch or warning

The "owner" on a 1990s record and the "seller" on today's agreement frequently do not match on paper — not because of fraud, but because of transliteration (Odia and Hindi names rendered differently into English across decades), initials versus expanded names, and inheritance that was never mutated into the records. Each mismatch has to be reconciled before it is safe to assume the seller is the owner.

Findings 4 and 5 — plot identity (44% discrepancy) and registered-transaction gaps (20%)

Plot, khata, and area figures disagreed across documents in 44% of cases (a Schedule-A that does not match the record-of-rights), and 20% had a gap in the registered-transaction history. These are the quieter defects — easy to miss by eye, decisive in a dispute.

If you are buying — the 5-point title check this data argues for

1. Trace the chain backward, not just the latest record — confirm each transfer links to the prior owner.
2. Pull a fresh Encumbrance Certificate and check for any charge and its release.
3. Reconcile the owner's name across every document, accounting for transliteration and inheritance.
4. Cross-check plot, khata, and area figures across the deed, the record-of-rights, and the map.
5. Confirm the registered-transaction history is continuous, with no unexplained gaps.

The honest limits of this data

The sample is small (59 complete checks) and self-selected; these are observations from a verification casebook, not a national survey. We will update the figures as the casebook grows. What the data does show, reliably, is that "the documents looked fine" is not the same as "the title is clean" — and the gap between those two is where buyers lose money.