To estimate the effect of selective sampling on first contact (FC) studies of the relation between migration and schizophrenia. We compared the FC method directly with a more inclusive longitudinal psychiatric register (LPR) method, by letting both methods estimate age and sex adjusted incidence rate ratios (IRR) in the population of The Hague aged 20-54 years, for the three largest migrant groups (first and second generation Caribbean, Turkish, and Moroccan) relative to the native Dutch population. Both methods found that the adjusted IRR was higher for migrants than for native Dutch [all migrants IRR = 1.70 (95% Cl 1.30-2.21) for the LPR method and 1.91 (95% Cl 1.15-3.25) for the FC]. The IRR for Moroccans was significantly lower in the LPR [IRR 2.69 (95% 2.10-3.41)] than in the FC study [4.81 (3.41-6.68)]. The FC method was relatively more inclusive for migrants presenting at earlier ages or with shorter durations of prior treatment (DPT) than the native Dutch. This resulted in differential sampling and artificially higher IRRs for Moroccan and, to a lesser extent, Turkish migrants. We confirm that the incidence of schizophrenia is raised twofold for migrants compared to nonmigrants. Using the LPR method, however, IRR estimates were less pronounced for most migrant groups than in a high quality FC study conducted in the same population. The FC method may overestimate the risk of schizophrenia for migrant groups who seek first mental health at a relatively younger age, or who present directly with schizophrenia.