Family history interviews are widely used in psychiatric research, as well as in genetic and twin studies, and provide a way to collect family history information quickly and economically. To obtain a valid assessment of family history, it is important to investigate which family member will be able to provide accurate information. Previous research shows that the validity of family history reporting can be influenced by characteristics of the informant, such as age, gender and personal history of psychiatric disorder. The aim of this study was to investigate the role of a subject's position in a pedigree on the validity of data collection. Family history data on diabetes and psychiatric disorders were collected in three generations of 33 families by interviewing both an index subject (3rd generation) and his or her mother (2nd generation). Mothers were shown to report higher rates of diabetes and psychiatric disorder in the family compared to the index subjects. There was no significant difference in the disease rate reported by male and female index subject. Mothers who experienced a depressive episode indicated significantly more family members as having a psychiatric disorder than mothers who never experienced such an episode. This could be explained by the presence of informant bias, but may also result from the fact that depression is a heritable disorder and is therefore actually more prevalent in these families. Our findings suggest that family interview data should be collected by interviewing subjects who have a central position in the pedigree and can therefore provide information on his/her own generation, the previous and the next. In addition, psychiatric status of the informant should be carefully addressed.