The paper describes recent experience in incorporating capacity constraints and crowding impacts in transit forecasting for the Los-Angeles County Metropolitan Transportation Authority (LACMTA).

The first aspect is to ensure feasibility of transit ridership forecast for each line and segment with respect to the total capacity constraint. The method to address infeasible volumes is to increase transit wait times at the corresponding boarding stations, i.e. use effective headways rather than schedule-based headways. This is based on the assumption that the riders will not always be able to board the first-arriving vehicle and will have to wait for the next vehicle. The second aspect is to account for crowding in the vehicle as a negative factor in the user perception of transit service quality. Based on a combined Revealed & Stated Preference survey a discrete choice model for transit mode & route choice was estimated with the following main results:

• Crowding is perceived by transit users as a negative factor expressed as a perceived weight on in-vehicle time that can reach a value of 1.6 or higher for extreme crowding levels.

• Crowding is more onerous for commuters to work experiencing it daily than for occasional trips for non-work purposes.

• Crowding is more onerous for older riders compared to younger riders.

• Crowding is perceived differentially by mode where bus has in general a greater (negative) in-vehicle time coefficient while commuter rail has a relative spike only at extreme crowding levels.

• The data suggested that all income groups are equally sensitive to crowding in terms of in-vehicle time coefficient differentiation by crowding levels. The difference between them is in the value of time that affects the cost coefficient rather than time coefficient.

• Crowding affects long trips more than short trips but it is well accounted by the length of the trip with the fixed crowding weight per minute.

The developed crowding functions and capacity constraints were incorporated in the operation travel model used by LACTMA. Application of the developed methods was very beneficial in practice and allowed for evaluation of transit projects in a more comprehensive way.