This project aims to understand how self-assessed health status relates to preferences for cycling infrastructure. An integrated latent class and latent variable choice model is fitted using responses to a stated preference experiment from a panel of New York City residents (N = 801). Estimates show that people with stated good physical health tend to have preference parameters similar to those of experienced cyclists. This result means that the provision of cycling infrastructure with the purpose of attracting non-cyclists also has the potential of attracting those with worse health outcomes. This result suggests a double benefit coming from car use reduction and lower health spending.