A hospital-based case-control study on soft tissue sarcomas (STS) was conducted in 1983-84 in Torino and in Padova (Italy). Cases (36 children with rhabdomyosarcoma (RMS) and 16 non RMS-STS) were compared to 326 controls. Histories of parental smoking habits and occupations, parental and children's exposure to ionizing radiation, children's diseases and some other variables were collected through interviews to the relatives attending the child in the hospital. A non statistically significant association was observed with both maternal age above 30 at child's birth (STS: OR = 1.5, C.I. = 0.8-2.9; RMS: OR = 1.9, C.I. = 0.9-4.0) and "in utero" exposure to diagnostic radiation (STS: OR = 1.9, C.I. = 0.5-6.5, based on 4 cases). No association was found with children's previous diseases. Paternal and maternal smoking habits were similar for RMS and STS cases and controls. Some positive associations with either maternal or paternal occupational histories were identified. They are difficult to interpret in view of the large number of comparisons and small absolute figures. They included maternal employment as medical doctor and nurse, farmer, textile worker and machine tool operator. An association was also observed with paternal occupation as butcher, building worker or employment in the production of domestic appliances. One case and no controls reported a maternal aunt affected by breast cancer.