"Business training is a widely used development tool, yet little is known about its impact. We study the effects of such a business training program held in Central America. To deal with endogenous selection into the training program, we use a regression discontinuity design, exploiting the fact that a fixed number of applicants are taken into the training program based on a pre-training score. Business training significantly increases the probability that an applicant to the workshop starts a business or expands an existing business. Results also suggest gender heterogeneity as well as the presence of financial constraints."