The ten richest and ten poorest municipalities of Sweden were investigated with respect to national statistics to assess the relationship between suicide incidence, discrete social psychological variables associated with welfare admittance among the elderly, and income of municipality. The relative frequency of suicide was 1.6 times greater for Swedes from the low-income municipalities than for those from the high-income ones. The group of municipalities with the highest suicide rate had a significantly higher proportion of older people in need of municipal social assistance at their homes and also a significantly higher proportion of elderly living in municipality-managed 'service-homes.' The findings replicate earlier investigations and suggest social psychological indicators denoting less favourable economic and social resources are also associated with both an increased suicide rate and a decreased county or municipal income. Some theoretical issues of the socioeconomic and of the external restraint hypotheses of the incidence of suicide, contradicted by the present findings as well as of Durkheim's hypothesis of social control are discussed. Further, we suggest the consideration of negative socioeconomic conditions as a risk factor amid psychiatric clinical assessments of risk for suicidal behaviour.