The article studies the demand for public transport in Colombian cities, identifying the main determinants explaining market shares for transit services. We used aggregated choice models to estimate overall price, income, speed and frequency elasticities for urban public transport demand in the country. The model successfully accounts for unobservable information about attributes of alternative public transport modes. The results suggest high demand elasticities to price and frequency, while demand responds moderately to speed. We also find a negative income elasticity, which classifies public transport as an inferior good. Results provide relevant information for the current discussion about new transport systems in cities of developing countries, with a deeper understanding of the elements to consider in the design of public transport policies. © 2020 Elsevier Ltd