Cycling behaviour in 17 countries across 6 continents: levels of cycling, who cycles, for what purpose, and how far?

International comparisons of cycling behaviour have typically been limited to high-income countries and often limited to the prevalence of cycling, with lack of discussions on demographic and trip characteristics. We used a combination of city, regional, and national travel surveys from 17 countries across the six continents, ranging from years 2009 through 2019. We present a descriptive analysis of cycling behaviour including level of cycling, trip purpose and distance, and user demographics, at the city-level for 35 major cities (>1 million population) and in urbanised areas nationwide for 11 countries. The Netherlands, Japan and Germany are among the highest cycling countries and their cities among the highest cycling cities. In cities and countries with high cycling levels, cycling rates tend to be more equal between work and non-work trips, whereas in geographies with low cycling levels, cycling to work is higher than cycling for other trips. In terms of cycling distance, patterns in high- and low-cycling geographies are more similar. We found a strong positive association between the level of cycling and women’s representation among cyclists. In almost all geographies with cycling mode share greater than 7% women made as many cycle trips as men, and sometimes even greater. The share of cycling trips by women is much lower in geographies with cycling mode shares less than 7%. Among the geographies with higher levels of cycling, children ( 60 years) remain underrepresented in all geographies but have relatively better representation where levels of cycling are high. In low-cycling settings, females are underrepresented across all the age groups, and more so when older than 16 years. With increasing level of cycling, representation of females improves across all the age groups, and most significantly among children and older adults. Clustering the cities and countries into homogeneous cycling typologies reveals that high cycling levels always coincide with high representation of females and good representations of all age groups. In low-cycling settings, it is the reverse. We recommend that evaluations of cycling policies include usage by gender and age groups as benchmarks in addition to overall use. To achieve representation across different age and gender groups, making neighbourhoods cycling friendly and developing safer routes to school, should be equally high on the agenda as cycling corridors that often cater to commuting traffic.