The list of surface route operating statistics for 2024 recently appeared on the TTC’s Planning webpage. This article consolidates data for the years 2019 (the last pre-pandemic year) to 2024.
Values included in the TTC files are:
- Weekday passenger count
- Weekday vehicle hours
- Weekday vehicle klometres
- AM peak vehicles
- PM peak vehicles
Derived values included in this article:
- Passengers (boardings) per vehicle hour
- Vehicle kilometres per hour
- Recovery rates relative to 2019
Pages for each set of stats are included in the detailed part of the article, and a PDF containing all tables is linked at the end.
In a future article, I will refresh a previously published table comparing service levels on routes in September 2025, when this information is available, with prepandemic conditions.
A basic message of both articles is that the TTC trumpets a return to former service levels, but the metric they use, vehicle hours, misrepresents the level of service provided. Buses and streetcars travel more slowly now than they did in 2019, and they have more generous recovery times at terminals. These combine to make a vehicle hour less productive in the amount of service it provides than in past years, and so 100% of former hours does not produce the same service. That is separate from other factors such as a decline in reliability, bunching and gaps which compound the less frequent scheduled service.
The problem is particularly bad on streetcar routes where a combination of factors including understaffing, budget constraints, and operating practices that slow service, lead to considerably less “recovery” to former service levels. In turn, this hurts service quality and blunts ridership recovery.
Crowding conditions are not addressed by these stats, and this is difficult to extract from TTC tracking data due to the coarse-grained nature of reported loads. TTC does not publish numeric vehicle loads, only broad light-medium-heavy loading indications. Although they report all-day ridership on each route, this is not broken down by time of day, location and direction, at least not for external consumption. TTC has crowding standards, but we do not know how well they meet them.
The number of peak vehicles is lower in 2024 than in 2019. This partly reflects limits on service growth, and partly the shift of demand into off-peak periods.
In the attempt to woo riders back onto the TTC, let alone to boost transit’s mode share for travel in line with City goals, the question of service level and quality is key. In theory, if demand actually sits below the historic level, then less service is needed to handle it. However, those who remember the condition of transit before 2020 will know that crowding was a pervasive problem and calls for better service were common. Only the March 2020 drop in demand saved the TTC from a capacity crisis.
Toronto must understand and commit the resources needed to achieve its transit goals. Just getting back to 2019 is no goal to aim at.
Continue reading




