An ordered-response model is used in this research to evaluate the factors that affect frequency of bicycle use for a person's commute to and from work. Data used were gathered during an original survey effort conducted over the Internet in 2002. Empirical results are presented, and policy implications of these results for urban planning are discussed. In addition, deterrents and facilitators of bicycle commuting as reported by respondents in the survey are descriptively analyzed. Several findings from this research contribute to the state of the knowledge in bicycle commuting. First, the availability of showers or clothing lockers at the workplace does not appear to inspire bicycle commuters to commute by bicycle more frequently. Second, using a bicycle for nonwork trip purposes increases an individual's frequency of commuting by bicycle to work. Other important results indicate that nonbicycle commuters either have misconceptions about the dangers of bicycling or else they lack convenient, safe route options for bicycling to work. Practitioners can use the ordered-response model to estimate an individual bicycle commuter's frequency of commuting by bicycle. The results can also help practitioners estimate the effects on nonmotorized mode share of programs that compete for funds to provide bicycling safety education, bicycle parking, and promotion of bicycling.