Background. Social differences in prostate cancer (PC) incidence and mortality might be related to testing for prostate-specific antigen (PSA). Although routine PSA screening is not recommended in Denmark, testing without clinical indication increased during the past decade. We evaluated associations between socio-demographic or clinical characteristics and PSA testing without clinical indication. Material and methods. In the Danish Diet, Cancer and Health Cohort, we identified 1051 men with PC diagnosed in 1993-2008. Diagnostic and clinical characteristics were obtained from medical records, and socio-demographic information was retrieved from administrative registers. We used general logistic regression analysis to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for associations between socio-demographic or clinical characteristics and PSA testing without clinical indication. Cox regression analysis was used to examine associations with mortality. Results. PSA testing without clinical indication was less likely among patients > 67 years (OR 0.7; 0.5-1.0). Men who were, PSA tested without clinical indication, were more likely to have vocational training (OR 1.8; 1.1-2.9) or higher education (OR 1.5; 0.9-2.5) and less likely to have advanced disease (OR 0.6; 0.4-0.9). PSA testing without clinical indication more often preceded therapy with curative intent (OR 1.8; 1.1-2.9) and less often palliative treatment (OR 0.6; 0.3-1.0). Men who were PSA tested without clinical indication had non-significantly lower overall and PC-specific mortality [hazard ratios 0.8 (0.5-1.2) and 0.6 (0.3-1.1), respectively]. Conclusions. PSA testing without clinical indication was associated with higher educational level. PC detected by PSA testing with no clinical indication was more often localized and treated with curative intent.