Pune Black Spot Tracker

What the police reported

2025, in Pune city police limits

Two details inside that report deserve to be read slowly. Of the men killed riding powered two-wheelers, 73% were aged 20–49 — working age, mid-commute. And of the women killed while walking, over 75% were above 50.¹ The people dying on Pune's roads are overwhelmingly the people who cannot afford a car and the people who are slowest across a junction.

Deaths did fall — 290 in 2025 against 341 in 2024.¹ That is real and worth saying. It is also why the black spot list matters more, not less: the remaining deaths are concentrated, and concentration is the condition under which fixing specific places works.

The problem underneath the problem

Nobody can tell you how many people died

This is not a gotcha. It is an ordinary data-governance failure, and it is the thing that makes every downstream claim — including mine — unverifiable. India's national road crash data comes from MoRTH, compiled from state police in a standard format, but it is published at state level and the widely-cited edition still reports 2022.³ There is no authoritative, machine-readable, city-level series for Pune that a citizen can download.

So the honest headline is not "290 people died." It is: 290 or 302, depending on which published source you read, and no public definition tells you why they differ.

The tracker

Five of twenty, and no status on any

Only five of the twenty black spots have been named in accessible reporting, with their crash counts. The other fifteen exist on a list I cannot read. Status columns stay empty until a public authority answers in writing.

Three of these five sit on named intercity corridors. That matters more than it looks: a junction on a state or national highway corridor may be owned by PWD or NHAI rather than the municipal corporation, which changes who is answerable and who holds the budget. Confirming ownership per location is the second question, not an afterthought.

Obstacles — the real ones

What will actually get in the way

• There is no primary document.The report was presented at a press event. I could not find a published PDF anywhere. Every figure on this page traces to news write-ups of it — including mine. That is a weak chain and it needs replacing with the document itself.
• Jurisdiction is undefined.City police limits, PMC, PCMC and Pune district are four different areas with four different numbers. Until a source says which one it counts, no comparison between years or cities means anything.
• "Black spot" is not one definition.Official highway criteria count crashes within a fixed radius over several years. A police list may use recent counts, local knowledge, or both. Two lists can disagree honestly.
• Hit-and-run at 54% breaks attribution.When over half of fatal crashes have no identified vehicle, cause analysis and enforcement-based fixes both lose their footing. Design fixes don't need a culprit — which is an argument for infrastructure over crackdowns.
• RTIs get deflected, not refused.The three standard moves are "information not maintained in this format", transfer to another department, and a partial answer that omits the number you asked for. All three are survivable if the question is written to be answerable.
• A tracker fixes nothing by itself.Budgets, road-owning agencies and elected pressure fix junctions. What a public tracker does is make the delay legible and dated, which is the raw material everyone else needs.

Realistic solutions

Small enough to actually happen

What one person can do

1. RTI for the report itselfAsk Pune City Police for a copy of the 2025 road safety report, the full list of all 20 black spots with locations and crash counts, and the definition of the area the fatality figure covers. This certainly exists, so "not maintained" is not available as an answer. ₹10, 30 days.
2. RTI for remedial works, per locationOnce the 20 are named: for each, the works sanctioned since 2023, amount, completion status, and current street lighting status. If the PIO isn't the right authority, §6(3) of the RTI Act requires them to transfer it within 5 days rather than reject it — so write it to be transferable.
3. Publish every reply verbatimScan, upload, date it. Including the unhelpful ones — especially the unhelpful ones. A refusal published next to the question is more persuasive than any argument.
4. First appeal, on timeNo reply in 30 days, or a non-answer, means a first appeal under §19(1) within 30 days, then the State Information Commission. Most people stop at step one; the appeal is where answers actually come from.

What the city could do that is genuinely cheap

• Publish the black spot list as open data.It already exists. Putting the 20 locations, their crash counts and their remediation status on a public page costs a morning of somebody's time and makes every other actor — journalists, councillors, residents — more effective at no further cost.
• Light the black spots first.37% of fatal crashes happen between 8pm and 3am. Lighting a named junction is weeks of work, not years, and does not require land acquisition or a flyover.
• Fix the crossings, not the drivers.90% of the dead were walking or on two wheels. Refuge islands, continuous footpaths and raised crossings protect exactly that group, and they keep working at 3am when no enforcement is present.
• State the jurisdiction on every published figure.One line of metadata. It is the difference between a number and a statistic.

Conclusion

The joke and the black spot are the same chowk

Pune adds roughly 1,356 vehicles a day to roads that are not growing.⁴ In 2025, somewhere between 290 and 302 people died on them, and nine in ten of those people were walking or riding a two-wheeler. The police know where it keeps happening — they published that there are twenty such places, and named five of them: Hadapsar, Kirloskar Pneumatic, Katraj Chowk, Perne Phata, Warje.

Those are the same junctions the whole city jokes about. The comedy is accurate; it is just missing the second half. What is absent is not awareness and not even, necessarily, action — it is any public record of whether action happened. That gap is small, specific, and closeable by asking in writing.

Method & sources

How to check me

Every figure on this page comes from a published source listed below. I have not computed, adjusted or estimated any road safety number. Where two sources disagree, both are shown rather than one being picked. The vehicle registration figure is my own arithmetic on an open dataset and the working is stated.

1. 1  Pune City Police road safety report, released 8 January 2026, covering 2025 — as reported by Punekar News. Secondary source. The primary document is what RTI #1 asks for.
2. 2  Free Press Journal, "Pune's deadly roads in 2025" — 302 deaths in 2025, 345 in 2024, 351 in 2023, 325 in 2022. Jurisdiction and methodology not stated in the article.
3. 3  Data For India on MoRTH road accident data — national and state level; the cited edition reports 2022. No city-level series for Pune.
4. 4  data.opencity.in, Pune vehicle registrations 2002–2017. Dataset row "Per year increased vehicle" as on 31.03.2017 = 494,795, which equals Total(2017) − Total(2016) as an independent check. 494,795 ÷ 365 = 1,356/day. Pune RTO region, 2016-17.
5. 5  RTI mechanics — ₹10 fee, 30-day PIO deadline, 48 hours where life or liberty is involved, first appeal under §19(1) within 30 days, State Information Commission within 90 days: RTI Wiki, Pune.