We aimed to investigate the role of perinatal determinants on the risk for germ-cell testicular cancer, with respect to the aetiological heterogeneity between seminomas and non-seminomas. A case-control study of 628 case patients with testicular cancer (308 seminomas and 320 non-seminomas) and 2309 individually matched controls was nested within a cohort of boys born from 1920 to 1980 in two Swedish regions (Uppsala-Orebro Health Care Region and Stockholm). Cases were diagnosed from 1958 to 1998 and were identified through the Swedish National Cancer Registry. Perinatal information on cases and controls was collected through charts available at maternity wards. Gestational duration, categorised in three categories (<37, 37-41, >41 weeks), was negatively associated with the risk for testicular cancer (P value for linear trend=0.008). A protective effect of long gestational duration and an increased risk for high birth weight were found for seminomas. Non-seminomas were associated with short gestational duration, particularly among those with low birth order (odds ratio: 3.02, 95% confidence intervals: 1.53-5.97) and high maternal age (odds ratio: 2.33, 95% confidence intervals: 1.19-4.55). No significant differences were found in tests for heterogeneity between the two histological groups. Our data support the hypothesis that intrauterine environment affects the risk for germ-cell testicular cancer. Seminomas and non-seminomas seemed to have similar risk patterns, although they are not entirely congruent.