This paper discusses methods for benchmarking vocational education and training colleges and presents results from a number of models. It is conceptually difficult to benchmark vocational colleges. The colleges typically offer a wide range of course programmes, and the students come from different socioeconomic backgrounds. We solve the comparability problem by focusing on effects in terms of retention and employment rates as opposed to the intermediate outcomes like grades. We neutralize cost differences using alternative cost measures. And we use detailed register data to account for student backgrounds. From a methodological point of view, we combine average methods (multilevel analysis) with frontier methods. We thus combine the key methods of school effectiveness research and school efficiency and productivity research. The analyses show that the efficiency of Danish vocational colleges varies considerably. Adopting best practices could lead to cost savings of between 9 and 33%.