[Response level of controls selected from the municipal registry]

The results related to the controls response rate in a case-control study are presented. The study was designed to learn about the relationship between the occupational exposure to toxic products in the parents and the development of acute leukemia in childhood. 128 cases were enrolled in the study and each case was matched by sex and age (less than 15 years) to a control. Controls were selected from the Municipal roll of the same city or town in which the case has been living during the diagnosis. We contacted the control's mother by phone or by personal visit. The response rate was 47.9% and near to a half of non responses was due to errors in the addresses registered in the Municipal roll. The validity of this source of controls in Spain can be questioned in the light of the high percentage of errors. We also discuss the potential selection bias introduced in the study due to the way we selected controls.