The work examines geographical variations of bicycle use in Trondheim municipality with the aim to create an understanding of what influences the use of bicycles and explain the geographical travel pattern for cycling. The work focuses on the how physical structures generates different transports effects. Where bicycle travels in Trondheim municipality are undertaken is described by using network analysis in ArcGIS, based on aggregated data from the travel habit survey (RVU) in 2009/2010. Trondheim municipality is divided into smaller geographical units (zoner) whose characteristics are described by 19 variables, categorized into 6 groups; Distance, Land use, Topography, Bicycle Facilities, Road characteristics and Demographic factors. By linear regression analysis in SPSS and ArcGIS Spatial Analyst, an explanatory model is elaborated. The variables distance to the regional center, workplace density, percentage separate bike paths and hilly terrain explains 78 percent of where cycle journeys are undertaken. Single variable with the highest explanatory power is degree of separate bicycle paths, which highlights the importance of infrastructure investment for bicyclists. Based on the result, recommendations are given on improvements of the bicycle routes in Trondheim, followed by a discussion of strategies for planning cycle routes.