Calgary Put a Bike-Lane Network on Trial—and Measured Everything

TL;DR;

• Calgary tested a 6.5-kilometre downtown network for 18 months against 82 performance measures.
• Weekday cycling on its routes tripled; downtown cordon counts rose 40% in one year.¹
• January counts averaged four times their pre-network level after the city committed to prompt snow clearance.¹
• Perceived safety improved and sidewalk riding fell, but one year of collision data could not prove a crash reduction.
• Council voted 10–4 to make the network permanent and fund further design improvements.²

A city made itself state the hypothesis

Most bike-lane debates begin with predictions. Ridership will soar. Traffic will seize up. Shops will lose customers. Crashes will fall—or migrate to intersections. Then a city builds one corridor, takes a few counts, and everyone returns to the argument they already preferred.

Calgary tried something more useful. Before opening its Centre City Cycle Track Network, the city specified what success and failure would look like. Its evaluation plan set targets for safety, bicycle volumes, driving time, unlawful riding, satisfaction, business activity, and rider demographics.³ The city then installed a connected network with temporary materials, measured it for 18 months, and brought the evidence back to council.

That process did not produce a perfectly controlled experiment. It did produce one of North America’s most unusually broad real-world evaluations of protected bike lanes. The findings are valuable precisely because some are strong, some are ambiguous, and some show how not to measure an effect.

Calgary’s larger lesson is not merely that bike lanes “worked.” It is that a city can put an entire network on trial, learn while it operates, and still make a decision without pretending every metric points in the same direction.

What Calgary actually built

Council approved the pilot in April 2014. On June 18, 2015, Calgary opened a 6.5-kilometre network through downtown and the Beltline, including protected routes on 5 Street S.W., 12 Avenue S., and 8/9 Avenue S., plus daytime cycling on the Stephen Avenue shared street. The project cost **C$5.45 million**, C$1.65 million below its C$7.1 million budget.⁴

The word network matters. Rather than test a lone strip of protection that ended before reaching useful destinations, Calgary connected east–west and north–south routes to the existing river pathway system. The city’s own counts later found the most riding where routes were close together and intersected, and the least where a track served fewer destinations. Riders were even detouring a block or two to reach the protected grid.¹

The project was also deliberately adjustable. Temporary curbs, posts, signs, and signals allowed staff to change loading, parking, sightlines, and intersections while people used the streets. Calgary logged 2,032 related service requests by November 2016—55% during construction and the first month—and made more than 100 operational or design changes.⁴

That is a different model of a pilot: not a miniature project designed to offend nobody, but a usable network treated as applied research.

The 82-measure scorecard

Calgary gathered baseline data in September 2014 and final data in September 2016 at ten locations, with six measurement periods and limited January collection to capture winter conditions. The tools included automated pavement counters, video and manual observation, police collision records, intercept surveys, GPS and stopwatch driving runs, and a third-party telephone survey of 1,102 Calgarians.¹⁴

The table below separates what Calgary measured from what the result can reasonably establish.

The city reported that 70% of its primary measures either met their target or moved toward it.⁴ That formulation is candid: several route-level volume and collision-rate targets were missed. A credible evaluation should make those misses visible rather than reduce dozens of measures to a pass/fail slogan.

The clearest result: people used the network

By November 20, 2016, automated counters at the middle of the three main corridors had recorded 1.2 million bicycle trips. Comparing September 2014 with September 2016, daily weekday cycling across the pilot network tripled. At representative locations, 16-hour counts rose from 630 to 2,040 on 5 Street, from 190 to 870 on 12 Avenue, and from 920 to 1,240 on 8 Avenue.¹

The annual downtown cordon count tells a related but broader story. Trips entering and leaving downtown rose from 12,304 in May 2015 to 17,193 in May 2016—Calgary’s largest recorded one-year increase, approximately 40%. Morning inbound bicycle mode share reached 3%, up from 1.9% in 2010.¹

Those figures should not be mashed into one claim. The tripling describes use on pilot routes; the 40% describes all bicycle crossings of the downtown cordon. Together they suggest both route concentration and a wider increase in cycling. They do not reveal exactly how many people were new riders, how many shifted from parallel streets, or what travel modes their bicycle trips replaced.

Still, route choice is part of infrastructure’s purpose. A protected track that draws riders away from an uncomfortable parallel street has succeeded even before it generates a wholly new trip.

Winter was an operations test, not a weather excuse

Calgary is a useful counterexample to the claim that protected lanes are wasted wherever snow falls. Comparing January 2015 with January 2016, the pilot found four times as many daily winter bicycle trips on average. Between November 2015 and the end of March 2016, the counters recorded more than 160,000 trips.¹

Concrete and paint did not cause that outcome alone. Calgary committed to clear the cycle tracks, like the pathways feeding downtown, within 24 hours after snowfall. The pilot therefore tested a service as much as a structure: a route that vanishes under snow is not year-round transportation.

Earlier Calgary research had already found a constituency for winter riding. In an intercept survey near freezing, 96% of frequent winter cyclists rode to work and 71% said they tolerated temperatures of −20°C or colder; better snow and gravel removal was their most requested improvement (Amiri and Sadeghpour 2015).⁵ The pilot then showed what happened when maintenance met that demand.

For practical cold-weather riding advice, see our guide to biking in cold weather. The policy point is simpler: winter suppresses cycling, but a city also chooses how much suppression to accept through its plowing standard.

Safety improved in experience before it was proven in crashes

The pilot made riders feel safer. In Calgary’s 2016 survey, 91% of people cycling on the routes reported feeling safe, and 77% said cycle tracks made downtown cycling better. Observed sidewalk riding collapsed from a network average of 16% to 2%, suggesting riders no longer felt as much need to improvise a refuge meant for pedestrians.⁴

The collision evidence was less conclusive. During one year of the pilot, police recorded 39 bicycle collisions, including 26 injury collisions and no fatalities, on the routes. Bicycle collision counts increased from before the pilot while bicycle traffic also increased. The city reported that motor-vehicle collision frequency decreased or held steady on each cycle-track corridor, and an external in-service road-safety review found no major safety issue, but it recommended minor adjustments.⁴⁶

Calgary’s evaluators explicitly warned that collision data vary sharply from year to year and are normally assessed over three to five years. Police records can also arrive late while court proceedings remain unresolved.¹ A one-year before-and-after comparison, on streets whose bicycle volumes changed dramatically, cannot settle whether risk per rider fell.

Longer-run research is encouraging but still uncertain. A 2026 difference-in-differences study covering cycling infrastructure installed in Calgary, Toronto, and Vancouver from 2011 through 2022 found Calgary cycle tracks associated with 2.1 fewer cyclist deaths or serious injuries per 10 kilometres, but the 95% confidence interval ran from 4.5 fewer to 0.3 more—meaning the Calgary estimate was compatible with no effect. The same study estimated a 207% increase in cyclist volume on sampled Calgary cycle tracks (Batomen et al. 2026).⁷

The honest conclusion is not “the pilot proved safety.” It is that usage and perceived safety improved immediately; observed severe-injury outcomes over a much longer period are consistent with a safety benefit, but still statistically imprecise. And because protection can fail at junctions, the result reinforces the need to keep improving the intersections where bike lanes too often end.

The riders changed, too

At three central count locations during six-hour weekday observation periods, the number of women riding rose from 200 in 2014 to 760 in 2016. Their share grew from 22% to 30%. Observed children increased from one to 32, or 0.1% to 1.3%. On a Saturday in August 2016, women were 33% of observed riders and people under 18 were 3.3%; the share over 65 was 1.1%.¹

These are not citywide demographic rates. Six-hour observations can be affected by weather, school schedules, work patterns, and which sites were selected. They also cannot tell us anyone’s motivation.

But the direction is meaningful. Before the pilot, Calgary’s downtown riding population was overwhelmingly adult and male—the profile expected when cycling requires mixing with traffic. The shift matches broader evidence that protection particularly matters to groups deterred by traffic stress, discussed in The Infrastructure That Brings Women Back to Bikes.

What happened to driving, walking, and parking?

The apocalyptic traffic forecast did not appear in the measured travel times. A full 5 Street trip took ten seconds longer in the afternoon peak. On 8 Avenue, the morning trip became 15 seconds faster and the afternoon result did not change. The 12 Avenue afternoon trip became four minutes faster, although Calgary attributed much of the baseline delay to nearby construction. Its morning trip did become 90 seconds slower, alongside an increase in average delay at 5 Street from 11 to 24 seconds per vehicle.⁴

These results need the same discipline as the collision results. Manual travel-time observations were collected on a single day in each evaluation period, downtown traffic declined generally, and construction obstructed some measurements. “No detected network-wide traffic disaster” is warranted. “Cycle tracks made traffic faster” is not.

Survey responses were mixed rather than catastrophic. Among people who had used the routes, 54% of drivers and 84% of pedestrians said their experience was the same or better; 77% of cyclists said cycling was better. Sixty-seven percent of drivers said the tracks did not change their route choice. Calgary also reported a net gain of 130 downtown parking spaces after creating new spaces elsewhere to offset project and routine inventory changes.¹⁴

The business numbers are a lesson in measurement

Calgary asked pedestrians how often they visited corridor businesses and how much they spent, while merchants estimated daily customer counts. Average reported visits stayed at 3.5 per week, but reported monthly spending fell from C$153 to C$131. Merchants’ estimated customers fell from 112 to 92 per day, although three-quarters said the tracks brought a similar number or more.¹

Those findings should neither be hidden nor treated as proof that the lanes caused a decline. The samples changed between waves, the measures were self-reported, there were no untreated comparison streets, and Calgary was experiencing a broader economic downturn with major Centre City job losses. City staff therefore classified economic vitality as a secondary measure and said the track’s influence was difficult to extract.⁴

A later scholarly review used Calgary as an example of why research design matters. Volker and Handy found the Calgary business result negative in direction, but noted that it lacked statistical testing, control sites, and adjustment for larger economic trends. Across the wider North American literature, studies with statistical testing found either positive or statistically insignificant effects on nearby retail and food-service businesses (Volker and Handy 2021).⁸

In other words, Calgary measured business concerns but did not identify a causal business effect. That distinction complements our broader article on the economics of cycling: a citywide benefit estimate and a corridor-level shop analysis answer different questions.

Why the experiment became permanent

In December 2016, Calgary City Council voted 10–4 to make the Centre City tracks permanent. It directed the C$1.65 million project savings toward permanent signals and continued changes to traffic operations, parking, and business loading.²

Council was not choosing a flawless scorecard. Only half of ten bicycle-volume count locations met their individual targets, the short-term cyclist collision target was not demonstrated, one peak-period route became 90 seconds slower, and the business evidence was unresolved. But the network carried more people, worked through winter, broadened participation, reduced sidewalk and wrong-way riding, attracted two-thirds public support, and imposed modest measured impacts on other modes.¹⁴

As of 2026, Calgary’s downtown cycle-track page still describes the pilot as the foundation of the network in use today and publishes links to public bicycle counters.⁹ Permanence did not freeze the pilot design in amber. It made durable upgrades and continued monitoring worth doing.

What another city should copy—and improve

Calgary’s process offers a practical blueprint:

1. Test a network, not an orphan segment. Connectivity made the routes useful and let the city observe a genuine route-choice response.
2. Publish targets before opening. Predetermined measures make it harder to cherry-pick a favorable statistic after political positions harden.
3. Use temporary materials for real learning. More than 100 adjustments turned public feedback and collision reviews into physical changes.
4. Fund operations from day one. Snow clearance was part of the infrastructure, not an optional maintenance detail.
5. Count who rides, not only how many. Gender and age observations tested whether the route appealed beyond existing confident riders.
6. Use controls and administrative data for economic claims. Merchant impressions and intercept spending surveys are useful signals, but tax receipts, employment data, comparison corridors, and multiple years would better isolate effects.
7. Keep measuring crashes after the vote. Eighteen months can uncover conflict locations; it cannot reliably estimate rare severe outcomes.

The best pilot is not a temporary lane waiting to be judged by anecdotes. It is a full-scale piece of public infrastructure with questions attached—and a plan to learn from every answer, including the inconvenient ones.

FAQ

Q1. Did cycling really triple during Calgary’s cycle-track pilot?
A. Yes. Weekday trips on pilot routes tripled between September 2014 and 2016; the separate downtown cordon count rose 40% in one year.¹

Q2. Did Calgary prove that its cycle tracks reduced crashes?
A. No. Perceived safety improved, but one year of variable post-opening collision data could not establish a reliable reduction in cyclist risk.¹⁴

Q3. How much did Calgary’s cycle tracks delay drivers?
A. Most sampled peak trips changed by seconds or not at all; the largest measured increase was 90 seconds, with important measurement limitations.⁴

Q4. Were Calgary’s protected bike lanes used in winter?
A. Yes. January counts averaged four times the baseline, with more than 160,000 trips recorded from November 2015 through March 2016.¹

Q5. Did the pilot hurt local businesses?
A. The surveys showed declines, but without control corridors or statistical modeling they could not separate the tracks from Calgary’s economic downturn.⁴⁸

References