Abstract
A direct ridership model (DRM) for predicting bus rapid transit (BRT) patronage in Southern California is estimated. Attributes of bus stops and their surroundings constitute the data observations of DRMs and enable a fairly fine-resolution analysis to be carried out on factors that influence ridership. The best-fitting DRM revealed that service frequency strongly influenced BRT patronage in Los Angeles County. High inter-modal connectivity, with both feeder bus routes and rail transit services, also significantly induced BRT travel. As population densities increased, so did BRT patronage, all else being equal. For exclusive-lane BRT services, high employment densities were also associated with high daily boardings. The strong statistical fit of the model bodes well for DRM as a platform for estimating BRT patronage in coming years.