Differences found in the incidence of suicide between the psychiatric catchment areas of the Karolinska Hospital in Stockholm were investigated in relation to health and socioeconomic indicators during the study period 1990-1994. The hypothesis of the study was that negative socioeconomic indicators and psychosocial and health indicators denoting less favourable socioeconomic status may negatively influence the suicide rate of the demographic units in this cross-sectional study. The incidence of suicide between the areas was significantly different and increasingly divergent in the last year of the study period. The area with a higher proportion of suicides had also an increased proportion of individuals who retired early, lower life expectancy at birth, higher non-employment, lesser income among the employed, less public expenditure for education, less proportion of home ownership, and a higher proportion of persons bound to one-room dwellings. Disregarding the influence of ethnicity (there were no statistically significant differences on immigrants' suicide between the areas) as well as in the availability of psychiatric care (assuming that similar quality of psychiatric care was provided by both sectors), or other demographic indices commonly shared by the areas, the possibility of strong effects of unfavourable health and socioeconomic indices appeared relevant for the explantation of an increased incidence of suicide. The findings provide new empirical contradiction to the socioeconomic hypothesis of the incidence of suicide, which postulated that populations with higher socioeconomic status may have increased suicide rates.