How safe is it to walk on India’s roads?

How many pedestrians are killed on India's roads?

The Ministry of Road Transport and Highways (MoRTH) reports 36,526 pedestrians killed in road crashes in 2024 in Road Accidents in India 2024. The figures come from state and union-territory police returns. Pedestrians were one in five of the 177,175 people recorded as killed on India's roads, close to 100 a day in a leap year. Only two-wheeler users formed a larger victim group.

A pedestrian is the person recorded as walking when they were killed. The label does not describe who caused the collision. It also does not tell us how dangerous a walk is compared with a ride or drive. Answering that needs deaths divided by the number or distance of walking journeys, alongside the same measure for other forms of travel. Road Accidents in India 2024 does not provide those denominators. The count still matters: walking appears in a large share of the road-death record even before we know the chance of harm on any one trip. This is an account of fatal crashes, so it cannot describe injuries, near misses, inaccessible streets or the fear that shapes where people walk. The broader road-death article places this toll beside the other road-user groups.

In June 2026, well after the deaths counted here, the Supreme Court recognised walking as a fundamental right. In Maniyar Iliyaz v. P. Ayyappan, the Court said that this right includes access to demarcated footpaths and takes priority over movement by motorised vehicles. The ruling followed the death of a five-year-old boy struck by a tanker while walking to school where there was neither a footpath nor a pedestrian crossing. Set beside the national death count, it sharpens the civic question: what space had been made for the people expected to walk?

Is the pedestrian toll falling?

No, at least on the latest comparison Road Accidents in India 2024 provides. Recorded pedestrian deaths rose from 35,221 in 2023 to 36,526 in 2024, an increase of 1,305 people or 3.7 per cent. Pedestrians made up 20.4 per cent of reported road deaths in 2023 and 20.6 per cent in 2024. The share moved little while the number of people lost rose.

Two years do not establish a persistent trend. Counts can move with travel volume, crash severity and reporting. A longer series would require joining annual reports carefully. MoRTH says it changed the format for identifying victim road users from 2019, and earlier reports classified deaths primarily by the impacting vehicleImpacting vehicleThe vehicle police recorded as the collision counterpart to a victim.This administrative label is not a court finding of fault.. Pre-2019 labels should not be spliced into this series as though they meant the same thing. This article stays with the 2024 victim data and the report's direct 2023 comparison.

How old were the pedestrians who died?

The largest published age band is 45–60, with 8,436 pedestrian deaths. Another 6,160 victims were 60 or older, and 1,888 were younger than 18. The age of 792 pedestrians was recorded as unknown, and the chart leaves that group visible. These counts sum to all 36,526 recorded pedestrian deaths.

The distribution resists a simple story that walking deaths belong to only one stage of life. Children, working-age adults and older people all appear. The report's age bands are broad, so we cannot see what happened at a particular birthday or estimate the risk of a school walk. Nor can we infer that a 60-year-old faces more danger per kilometre walked simply because an age band has more recorded deaths. Groups differ in size and in how much they walk. The next comparison asks a narrower question: among people killed on roads at each age, how many were pedestrians?

Are older road victims more often pedestrians?

Yes, as a share of recorded road deaths within the same age band. Of the 14,767 people aged 60 and over killed in road crashes in 2024, 6,160 were pedestrians, or 41.7 per cent. Among victims aged 18–25, 4,333 of 34,154 were pedestrians, or 12.7 per cent. The bars show that walking takes a different place in the victim mix at different ages.

The calculation matters. For each age band, pedestrian deaths were divided by all road deaths in that band and then multiplied by 100. Each age band has its own total of road deaths. A larger percentage among older victims could reflect how people travel, what happens when they are struck, or both. It is not the probability that an older person dies while walking. That would require counts of older people walking, their journeys or distance, and comparable exposure for other ages. The report supplies none of those.

How does the pedestrian toll differ for women and men?

The police record lists 29,055 male and 7,471 female pedestrian victims in 2024. Men account for 79.5 per cent of pedestrian deaths and most recorded road deaths overall. Looking inside each group reveals another part of the story: pedestrians made up 29.6 per cent of female road victims against 19.1 per cent of male road victims. Each share uses all road deaths recorded for that sex, not the number of people walking. MoRTH's Table 4.3 lists 25,225 female and 151,950 male road deaths; its pedestrian annexure lists the pedestrian deaths within those totals. The pattern holds in every published age band: among road victims aged 60 or older, 56.7 per cent of women and 37.9 per cent of men were pedestrians.

This is a difference in the mix of people killed on roads. It does not show that a walk is more dangerous for a woman than for a man. We do not know how many trips each group made, when and where they walked, or how often they crossed high-traffic roads. The published table has only male and female categories, so it tells us nothing about people outside them. The victim's recorded sex is not a crash-cause field either. The next table changes perspective, from the person who died to the vehicle police recorded in the collision.

What vehicles were recorded in pedestrian deaths?

Two-wheelers were the largest named impacting-vehicle group, recorded in 10,378 pedestrian deaths. Cars, taxis, vans and light motor vehicles appear in 9,302; trucks and lorries in 5,165. The “Others” category holds 7,857 deaths, more than one in five pedestrian victims. It mixes vehicle and other classifications behind a single label, so it is a substantial blind spot.

MoRTH calls the recorded counterpart a “crime vehicle” in its table. The term is administrative. A row saying a pedestrian was struck by a two-wheeler carries no legal verdict about the rider. The national table does not record the street layout, crossing, speed or sequence of actions. The bars also do not rank how dangerous each vehicle is per kilometre driven. They help identify which encounters need closer local study. A useful investigation would connect the crash record to the actual place where people crossed and vehicles moved.

Where are the largest pedestrian death totals?

Tamil Nadu reports 4,712 pedestrian deaths in 2024, followed by Bihar with 4,149, Maharashtra with 3,344 and West Bengal with 3,241. The ten states shown in the chart account for 27,354 deaths, 74.9 per cent of the national pedestrian total. The remaining 9,172 deaths are spread across other states and union territories.

These are counts of recorded loss, useful for deciding where more detailed reporting and investigation are needed. They do not rank how safe it is to walk. States differ in population, how much people walk, road networks, traffic and completeness of police reporting. Even a per-resident rate would miss differences in walking exposure. To compare the danger of a walking journey across states, we would need a consistent record of walking trips and pedestrian deaths. The national report provides the second half of that calculation, imperfectly, and not the first.

Among the highest-count states, how much of the road toll involves pedestrians?

The raw totals tell us where many pedestrians died. A second calculation asks what share of each state's recorded road deaths involved a pedestrian. Among the ten states in the previous chart, Bihar recorded 4,149 pedestrian deaths out of 9,347 road deaths, or 44.4 per cent. Uttar Pradesh recorded 2,131 out of 24,118, or 8.8 per cent. Tamil Nadu's 4,712 out of 18,449 was 25.5 per cent. The contrast shows that a large overall road toll need not have the same mix of victims everywhere.

These ten states were chosen because they had the largest pedestrian counts, not because they had the highest percentages. A higher share can reflect differences in how people travel, how crashes are recorded, or both. It is not the risk of a walking trip. West Bengal's displayed share also needs extra care: MoRTH says it recast that state's data using aggregate electronic accident-reporting figures because the return was not in the prescribed format. The chart is a prompt for local investigation, not a league table of safe places to walk.

Is this only a national-highway problem?

No. MoRTH records 11,386 pedestrian deaths on national highways in 2024, 31.2 per cent of the country's pedestrian total. Subtracting that count from 36,526 leaves 25,140 on all other roads combined. That second bar includes state highways and other roads; the source table does not break the pedestrian remainder into those categories here.

The larger off-highway count matters because a highway-only response would miss most recorded pedestrian deaths. It does not prove that another road type is more dangerous for a person walking. Highways and local roads carry very different volumes of traffic and pedestrian movement, and the report gives no common exposure measure to divide the deaths by. The split points towards a broad street-and-road question: people walking need protection in many settings, and the national total alone does not tell us which crossing or stretch should be fixed first.

What does the right to walk require?

The Supreme Court judgment identifies a duty that goes with the right. Where there is a road, urban development authorities, municipal bodies and panchayats must endeavour to demarcate, build, maintain and protect footpaths and other pedestrian infrastructure. The Court said people can seek legal remedies when that duty is breached, and it called for a dedicated legal framework. The judgment itself did not establish a new regulator or measure how many Indian roads have usable footpaths.

The national tables in Road Accidents in India 2024 do not identify a single cause of pedestrian deaths. They do not tell us whether each victim had a usable footpath or where the nearest safe crossing was. Nor do they give a reliable national measure of vehicle speed at the crash, or tell us whether a later hospital death was linked back to it. Reading the impacting-vehicle row as a verdict on behaviour would skip the place where the collision happened. The 2024 figures describe deaths recorded before the 2026 ruling, so they cannot test whether authorities have since complied with it.

The World Health Organization (WHO) pedestrian-safety manual describes the evidence behind separating pedestrians from traffic, reducing conflict at intersections and managing speed through road design, enforcement and other measures. Those are general findings. They do not measure India's 2024 streets, and they do not prove that a named measure would have prevented every death in this report. For an Indian city or district, first document the footpaths and crossings people can actually use. Then connect crash records to road layout, pedestrian movement, traffic speed and care after a crash. A national annual PDF can locate the burden. It cannot tell us which street has fulfilled the right to walk.

How should you read these figures?

The death counts come from the Ministry of Road Transport and Highways' Road Accidents in India 2024, assembled from state and union-territory police returns. Table 4.4 supplies the 2023–24 pedestrian totals and road-user comparison. Annexure 33 supplies pedestrian age and sex; Tables 4.2 and 4.3 supply all road deaths by age and sex for the within-group percentages. Table 4.5 classifies vehicles recorded in pedestrian collisions. Annexure 29(a) gives pedestrian state totals; Table 5.6 gives all road deaths in those states. Table 2.11 supplies pedestrian deaths on national highways. We used the original PDF, checked every plotted observation, and reconciled age, sex and state totals. The scanned annexures were read with a model only as a transcription aid, and the displayed cells were checked against the original page images.

The legal account comes from the Supreme Court's 19 June 2026 judgment in Maniyar Iliyaz v. P. Ayyappan, especially its conclusions on pages 12–13. It establishes a right and corresponding duties. It is not a source for the 2024 death counts, the prevalence of footpaths or the effect of a safety measure.

Shares of all 2024 road deaths divide by 177,175. Pedestrian-only shares divide by 36,526. The age-composition chart divides pedestrian deaths in each age band by all road deaths in that same band. The selected-state chart divides pedestrian deaths by all road deaths in each of the ten displayed states. West Bengal’s reporting was recast from electronic aggregates, which limits direct comparison. The sex-composition chart divides male or female pedestrian deaths by all road deaths of that sex. The age-and-sex comparison uses the same approach. “Other roads combined” is 36,526 minus 11,386 national-highway pedestrian deaths. None of these calculations gives deaths per walking journey or per kilometre walked.

The report's rural–urban, junction, weather and time-of-day tables cover road crashes generally. They do not identify the pedestrians killed in each category, so they cannot tell us where or when pedestrian deaths occurred. The 2023–24 comparison is short because victim classification changed in 2019. WHO's manual gives general evidence on pedestrian protection, but no India-specific measure of how much people walk. MoRTH says it compiled the national report from consolidated police returns, with West Bengal’s aggregate electronic data used for validation and recasting. Police reporting can miss deaths, especially when a later hospital death is not linked back to the crash record.