what is known already: Some associations between these outcomes have been shown in studies focusing on individuals and mainly in Nordic European countries. These associations, together with histological evidence of a dysgenesis pattern in testicular tissue specimens, have generated the concept of the existence of a 'testicular dysgenesis syndrome' originating in utero. study design, size, duration: This is a geographical correlation study using cancer, malformations rates and sperm quality data collected between the years 1998 and 2005.participants/materials, setting, methods: Incidence rates of testicular cancer were extracted from International Agency for Research on Cancer registries and Globocan, while cryptorchidism and hypospadias prevalence rates were obtained from EUROCAT and International Clearinghouse for Birth Defects Surveillance and Research registries. Sperm concentration data were extracted from recent studies using standardized methodology. A total of 39 registries and 9 sperm studies were selected. Non-parametric Spearman correlation tests were used to test the association between these four disorders. Correlations were computed for all registries together, for registries with high-quality matching coverage only and by continents. Sensitivity analyses were also conducted using data from prospective clinical studies to take into account potential bias related mainly to ascertainment of malformation rates.main results and the role of chance: We found positive correlations between testicular cancer and hypospadias (r ¼ 0.32, P ¼ 0.05) and between hypospadias and cryptorchidism (r ¼ 0.70, P ¼ 0.008). Stronger correlations were observed when using registries with high-quality matching coverage. Among these registries, differences between Europe and the rest of the world appeared (the positive correlation between testicular cancer and cryptorchidism was stronger outside Europe, r ¼ 0.83, P ¼ 0.01 compared with 0.40, P ¼ 0.60 for European registries). A negative correlation between testicular cancer and sperm concentration was observed (r ¼ 20.88, P ¼ 0.002). These correlations support our initial hypothesis but remain only suggestive due to the intrinsic limitations in the study design (i.e. geographical correlation study) and do not allow causal inference.limitations, reasons for caution: Differences in the ascertainment of malformations rates (definition, length of follow-up) make the international comparison difficult. The small number of registries for some conditions (cryptorchidism) or of studies (for sperm quality) and the absence of information about major risk factors such as ethnicity and socioeconomic status in the registries are also limitations.wider implications of the findings: Our findings are in agreement with results of studies focusing on individuals and suggest that shared risk factors are present in the populations studied. † These authors contributed equally to this work.