Is there any more recent data on how many pedestrians are struck by right-turning drivers? vs By left-turning drivers?
I recall data from back when the program started, just wondering if there is anything more recent.
Intuitively, the left turns are riskier because drivers are more likely to leadfoot around a left turn, due to oncoming cars and impatience around bad decisions with too-short gaps in same, and because they have room to accelerate during the turn. The saving grace of right turns from a stop is they are slower speed and less dependent on avoiding oncoming cars.
- Paul
No, very annoyingly, there isn't with publicly available data. Maybe someone on the Vision Zero team would have access to a different dataset but what is published on the Toronto Police Service website is far from quality data.
In February 2025 I emailed TPS asking when data would get updated for the Pedestrian KSI and was told maybe April and they were waiting on Vision Zero at the City. April rolled right by and nothing, so in October I contacted Vision Zero, and this was their reply;
"The City of Toronto Data & Analytics team is currently in discussions with the Toronto Police on coordinating the transfer of the Killed or Seriously injured dataset publishing authority to the City of Toronto. I do not currently have a timeline as this is a fairly large undertaking."
However, currently a couple sets of data capture some of what you're asking:
Traffic Collisions Open Data which has all collisions by all modes in the City of Toronto since 2014. 19,180 collisions involve at least one pedestrian. Seems to get updated somewhat frequently, every couple months and last update was November 2025. The biggest issue with it is incomplete or missing information. The schema includes, among other properties, date, hour, latitude, longitude, as well as properties to indicate whether an automobile or pedestrian was involved. But it doesn't include fault, actions, or position on the road. Of the 19,180 records involving a pedestrian, 1,494 don't have a lat and long. For the entire 772,516 records, 126,069 don't have a location! A big gap with this data is that each collision only has one record, but doesn't indicate how many people or parties were involved or what the injuries were.
There's another source,
Pedestrian KSI which has records going back to 2006, but the most recent record is from December 29 2023. From 2006 - 2023 there are 3,275 records of pedestrians involved in collisions (all modes). The schema includes "manoeuvre", with values including "Turning right" and "Turning left", but the critical piece missing is that this data is only capturing pedestrians who were killed or seriously injured. It does categorize injuries into None, Minimal, Minor, Major, Fatal, and <blank>, but the data very clearly does not include the many incidents involving pedestrians, by which I mean even if you read about a collision reported by the police resulting in a seriously injured pedestrian, it's not necessarily in the data. There isn't a definition about what qualifies for these injuries either. I recall Jess Spieker of Friends and Families for Safe Streets saying that her injury -- broken spine and a brain injury -- wasn't classified by police as "major".
Whether because police aren't providing complete information, or that not all incidents are logged, I am skeptical that a true picture of the pedestrian safety landscape can be painted with this data. In theory it should be very easy to compare data from this data to that of Total Collisions, but it isn't. They each have different unique identifiers and also event ids, so you can't even take a record of a collision in one table and easily find it in the other.
Comparing the two datasets in similar timeframes, 2014-2023, the number of collisions involving pedestrians are wildly different. Total Collisions has 19,180 collisions that involved at least one pedestrian (are they injured, a witness? I don't know), while Pedetrian KSI only has 1,623.
In July 2025 CAA released results from a near miss safety study they commissioned, though it was done in numerous Canadian cities and not just Toronto. I found the results are shocking, beacuse it captured near misses, and these happen all the time, and right turn conflicts happen
a lot more than what collisions resulting in injuries might lead one to believe (Toronto numbers below). I wrote a summary about it in
this comment.
Here's a link to the study:
https://www.caa.ca/app/uploads/2025/06/FINAL-REPORT_CAA-Intersection-Safety-Study_English.pdf
One graph from their results:
In any case since I have a database open here's some numbers:
Collisions from 2006 to 2023 in Pedestrian KSI
Turning right: 255
Injuries
None: 249
Minimal: 4
Minor: 2
Turning left: 816
Injuries
None: 771
Minor: 18
Minimal: 14
Major: 11
<blank>: 2
Within the Pedestrian KSI data, here's the breakdown of all injuries (2006-2023)
Minimal: 176
Minor: 222
Fatal: 546
<blank>: 567
Major: 2,626
None: 3,551
www.toronto.ca
