The paper describes problems and advantages in an international cooperative study of a case-control design, aimed at investigating the possible association between exogenous hormones and hypospadias. The varying degrees of the ascertainment of specific exposures and risk factors, in spite of the use of a standardized questionnaire is illustrated. Definite support for the existence of recall or interviewer bias is presented. On the other hand, the multipopulation design offers possibilities to make use of the diversity of the populations: differences in reproductive patterns and in specific exposures such as drug use, smoking and maternal occupation.