Planners have developed an appetite for complex and multifarious models to match the intricacy of the questions being asked. For the most part, these new models, such as the currently popular activity-based modeling framework, are seamlessly backward compatible with all the previous questions that planners still ask from time to time. However, a seemingly simple question was recently raised in San Francisco, California, and the San Francisco County Transportation Authority's advanced, state-of-the-art activity-based model SF-CHAMP (San Francisco Chained Activity Modeling Process) was not equipped to handle it gracefully. This paper documents a methodology used to create a schedule of the auto trip rates from the SF-CHAMP activity-based travel demand model. The linear regression methodology uses outputs from the SF-CHAMP model along with simple accessibility variables to account for the wide variations in vehicle trip rates across the city and to provide a nexus between trip generation rates and the context of their origins or destinations. This approach combines the desired simplicity of auto trip generation rates in the local context of San Francisco and sensitivity to a set of accessibility variables comprehensible to humans.