The methodological design, characteristics and fieldwork stage of a multilevel research study on the impact of the environmental characteristics on mental health in an autochthonous and immigrant population are described in this paper. Individual data were obtained using a core questionnaire 40 minutes length from home interviews of Spanish and Ecuadorian adults from September 2006 to January 2007. A random sample of 1186 people aged 18-55, with equal distribution of gender and nationality was obtained from Civil Registers of 33 areas (municipalities or neighbourhoods) of Madrid, Alicante, Almeria and Murcia, chosen by ethnic density and socioeconomic criteria. Previously, a pilot study was carried out. Socioeconomic indicators of neighbourhoods and selected communities were obtained from Municipal Registers and other secondary sources. Finally, 1144 people were interviewed (96%). Each person was contacted at home at two different times. The global response rate was 61%, higher among Ecuadorians (69%), who presented more problems of localisation (34%). Analyzing methods and fieldwork process the conclusion is that sample strategies for this type of population studies should be evaluated using feasibility criteria given time and money constraints, against the need to obtain representative samples of the target populations. There were serious shortcomings in the availability of social integration indicators at the neighbourhood level.