The current study is designed to examine long-term trends by histologic types of testis cancer in Connecticut. A regression model was used to identify age, period, or cohort as determinants of the time-trend on histologic types of testis cancer. The results from this descriptive epidemiologic study show that the overall age-adjusted incidence rate of testis cancer has increased 3.5-fold in Connecticut during the past nearly 60 years of cancer registration. The rates for seminoma and nonseminoma have been increasing since the mid-1950s and increase in a similar manner for those aged I5 to 49. The largest increase was observed in the age groups 20 to 44 for seminoma and 15 to 34 for non-seminoma. The observed increase was limited to whites. The results from age-period-cohort modeling suggest that the observed increase in seminoma before 1950s could be largely attributable to a period effect, while the increase for cohorts born after about I 9 I0 both for seminoma and for non-seminoma are mainly explained by a strong birthcohort effect. Therefore, the observed increase in germ-cell testis cancer in this population is likely to continue in the coming years. Thus far, the proposed hypotheses, such as exposure to DES in utero, earlier lifetime exposures to viruses, trauma or unusual amounts of heat to the testis, cannot adequately explain the observed incidence patterns of testis cancer. Analytical epidemiologic studies with large sample size are urgently needed to examine the risk factors responsible for the increase.