In four previous case-control studies we have shown an association between exposure to phenoxyacetic acids or chlorophenols and soft-tissue sarcoma (STS). These chemicals are contaminated by dioxins and dibenzofurans, some of which are known carcinogens in experimental animals. In a meta-analysis of our four studies we studied such exposure including the relation to tumour localization and histopathological type of STS. In total 434 histopathologically verified STS cases and 948 controls were included. Anatomical tumour localization was assessed by scrutinizing medical records for all cases. In the analysis stratification was made for age, vital status and study. Significantly increased risks were obtained for exposure to phenoxyacetic acids and chlorophenols, odds ratios 2.7 and 3.3, 95% confidence intervals 1.9-4.7 and 1.8-6.1, respectively. The increased risk was independent of histopathological subtype and anatomical site although especially high odds ratios were obtained for body sites close to areas where dermal contact may occur in sprayers. Phenoxyacetic acids and chlorophenols with their contaminating dioxins and dibenzofurans should be regarded as carcinogenic for STS regardless of tumour localization and histopathological type.