897 patients received radiotherapy in Scotland for tumours of the testis between 1950-1969. 299 patients who died within two years of treatment, and a further 51 patients with inadequate follow-up data, were excluded from a survey of the incidence of second tumours. 547 patients with a mean follow-up of 15.4 years were included in the study. Fifty-one second and six third malignancies were found. The observed incidence is significantly higher than expected (ratio 1.87; p less than 0.001), both in the irradiated sites (ratio 1.94; p less than 0.05) and in unirradiated sites (ratio 1.99; p less than 0.01). These patients have a higher incidence of second testicular tumours (ratio 23.1; p less than 0.001). Analysis of the incidence of new malignancies in successive quinquennia after irradiation shows a significantly higher incidence in two periods. Within five years the greater incidence (p less than 0.05) is due mainly to the occurrence of second testicular tumours. In the period 15-19 years after irradiation the higher incidence (p less than 0.01) is accounted for by tumours arising in the urinary and gastro-intestinal tracts. The increased risk of developing a second cancer is low and there was no observed increase in radiation-related leukaemia.