The bicycle is an efficient way to travel. There are individual and population-level health and wellbeing benefits that arise when more people cycle. However, cycling is rare in England and Wales and commuter cyclists are disproportionately likely to be male and middle aged (35 to 49). Society therefore misses out on the wider benefits of higher cycling levels, and women and certain age groups miss out on the individual level benefits. This thesis uses geospatial analysis to examine cycling behaviours at multiple scales, seeking to understand the interactions between demographics and causal factors of commuter cycling mode share. It also examines the influence of vehicular traffic in detail and considers what actions local authorities might take to increase cycling levels. Using both aggregate (area-based) and network (route based) modelling approaches, it identifies that the most important factors influencing cycling behaviours are hilliness, traffic, wealth, temperature and population density. Whilst these and other factors differ somewhat in their relative importance between demographic groups, differences in cycling rates are best explained by group-specific responses to the combined influence of all factors - the relative utility of cycling. On average, women and older (> 49) or younger (< 35) commuters require a higher level of utility before they start cycling to work. Findings also show how the different traffic characteristics of a commuter's route to work - such as vehicle speed, volume and direction - have distinct individual and combined influences on cycling propensity. Vehicle speeds are shown to be the most influential traffic characteristic. Policy should work towards making urban areas compact, dense and traffic free, with vehicle speeds under 30kph and with suitable levels of cycling infrastructure along key corridors to work. Urban form should be designed primarily with female cyclists in mind and male cyclists will benefit accordingly.