Background. Several studies during recent years have reported an increased occurrence of schizophrenia in selected immigrant groups. However, difficulties in establishing the population denominator for immigrant rates of mental disorder, selective referral for treatment, and other problems may have influenced such results. The current study aims at testing empirically the influence of some of these methodological problems.

Methods. Using nationwide case register data from Denmark, the diagnosis specific first-admission rates were compared between persons born in Denmark and other countries respectively. A case–control design was used to compare schizophrenia risk between different immigrant groups.

Results. Incidence rates of mental disorders among non-Danish residents calculated from admission data showed increased rates of schizophrenia (RR 1·7) and non-affective functional psychoses (RR 1·9). A case–control analysis utilizing non-psychotic admissions as control for schizophrenic admissions yielded essentially the same result, thus excluding selective referral as the sole explanation of the increased schizophrenia rate. However, this was almost exclusively due to increased rates in individuals born in countries neighbouring on Denmark.

Conclusions. While selective risk factors may be operating in various groups of immigrants, caution should be warranted in the interpretation of immigrant studies as large portions of transient visitors may obscure actual rates of mental disorders.