<b>Background:</b> The selection and the professional training for a resident in internal medicine requires a great investment of time, training effort and money. Drop-outs are therefore considered a failure of the selection procedure. To evaluate our selection model and outcome of the training programme, we determined the drop-out rate at the Leiden University Medical Centre with its affiliated hospitals.
<b>Methods: </b>Data were collected from all files that have been kept from 1988 onwards of all internists trained and registered in Leiden. These files contained the application forms, the assessments of the trainers and references and the specifications of the programme. Also the first employment as board-certified internist is registered.
<b>Results:</b> The drop-out percentage of the training programme was 8.5%. The drop-outs did not differ in study characteris-tics from those who successfully completed the programme. The reports from the training team members showed that the drop-outs were as suitable and motivated as the other candi-dates. During the training programme 8.5% of the residents moved to another university to complete their training as an internist. All board-certified internists who graduated at the Leiden University Medical Centre found a job.
<b>Conclusions:</b> The efficacy of the selection procedure for trainee internists is more than 90%. There are no studies in the literature for a comparison of data.