This research, which forms part of a larger study on Mental disorder and use of public sector psychiatric services among foreign residents in the canton of Ticino, focuses on the relationship between migration and health.Aim: The aim is to describe adult users of the public psychiatric services in three Swiss regions (Ticino, Zurich and Vaud), identifying possible differences linked to nationality, in relation to socio-demographic and clinical characteristics and contact with the services. Method: The patients involved in the study were all those between 20 and 49 years of age who had contacted the out-patient or in-patient psychiatric services in the regions concerned at least once during 1995 - in total 6794 in-patients and 13199 out-patients. For each of five nationality groups (Swiss, Italian, Iberian, ex-Yugoslavian and Other') the annual rate of first admission to psychiatric hospital was calculated as were the annual rates of in-patients and out-patients prevalence. Annual rates of in-patients and out-patients prevalence were also calculated for each diagnostic category, defined according to the International Classification of Diseases (ICD-10). Results and conclusions: The study confirmed the difficulty of describing similar patterns across the regions, due to a number of factors (problems of data standardisation, different migration and integration histories in the different host regions, regional differences in the psychiatric services offered: characteristics particular to the public psychiatric service, the importance of private psychiatric services, the existence of a health service network specifically for immigrants, etc.). Despite these difficulties, three different groups emerged in terms of in-patient and out-patient prevalence rates and, particularly, rates of first hospital admission: non-Italian and non-Iberian foreigners, who had the highest rates, Italians and Iberians, who had the lowest rates, and Swiss, who were generally in-between. The lower rates of the Italians and Iberians might be linked to a positive selection effect (healthy migrant effect), whereas for certain individuals in the Other' nationality group more complex factors could play an aggravating role (greater difficulty of integration, precarious status, the fact of having lived through particularly traumatic experiences in their countries of origin). As regards diagnosis, among out-patients those individuals originating in former Yugoslavia and 'Other' countries showed the highest rates of neurotic, stress-related and somatoform disorders (F4) and affective disorders (F3) (in certain cases roughly twice the rate shown by the Swiss, Italians and Iberians), which could be attributable to post-traumatic stress or to adjustment disorders, probably linked to factors such as type of migration, forced family separation, dramatic life events, language-related social and cultural problems. In contrast, no particularly important nationality-related differences were evident in hospital diagnostic rates. Some differences emerged with regard to reasons for out-patient referrals and to the number and duration of hospitalisations, which were generally higher among the Swiss and Italians; these data could be indicators of shorter psychiatric histories, of emergency first hospitalisations with no after-care follow-up, or of early interruption of treatment among non-Italian foreign nationals.