Recent transport policy favors cycling, yet it remains to be seen, in a car-dependent society, whether individuals will increase the amount they cycle. With Edinburgh, Scotland, as a case study, discrete choice models, based on random utility theory and using stated preference data, were developed to determine the propensity to cycle and provide an estimate of cycling demand. The data are from a survey of 997 households in West Edinburgh. Respondents in employment or education were split into population segments according to their propensity to cycle. The survey includes a cycling-based stated preference experiment for those traveling to work or study by car, by bus, or on foot. The study is novel in considering the population as a whole, instead of focusing on cyclists, with the modeling results split by population segments. Model estimation shows that cyclist facilities, primarily at the destination but also en route, largely determine the propensity to cycle to work or study.