In this study, a context-dependent discrete choice model to investigate cyclist’s preferences on link-level cycling infrastructure design features is to investigate the extent the preferences on cycling infrastructure design is affected by the characteristics of a road segment. Stated preference experiment utilizing visual representation of the alternatives were conducted. The model estimation indicates that cyclists attach different values on certain design features of a cycling infrastructure under different urban contexts. Nevertheless, there are significant heterogeneity on the preferences across respondents which still cannot be explained by the variation of contexts. Segmenting the data based on personal characteristic attributes, we found that differences in traveler’s personal characteristics are associated with certain differences in valuation of design features and also in sensitivity toward context effects.