In a case-control study on 43 patients operated for invasive breast cancer (cases) and 35 patients operated for benign breast disease (controls) adipose tissue concentrations of polychlorinated biphenyls (PCBs), 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene (DDE) and hexachlorobenzene (HCB) were investigated. Approximately 10 g of breast tissue free from tumour was taken and frozen until analysis. No significant difference for the sum of non co-planar PCBs or DDE was found between cases and controls. For postmenopausal women the odds ratio (OR) was increased for co-planar PCB #77 > 4.5 pg/g lipid (OR = 5.8, 95% confidence interval (CI) = 0.8-42), PCB #126 > 145 pg/g lipid (OR = 2.2, 95% CI = 0.2-18), PCB #169 > 90 pg/g lipid (OR = 7.8, 95% CI = 0.6-96), and for HCB > 40 ng/g lipid (OR = 1.9, 95% CI = 0.4-7.2) adjusted for age and parity. The risk increased further for postmenopausal women with oestrogen receptor positive tumours yielding for PCB #77 adjusted OR 33 (95% CI 1.8-588), PCB #126 OR not calculable (no unexposed cases), PCB #169 OR 8.6 (95% CI 0.5-136) and hexachlorobenzene OR 7.1 (95% CI 1.1-45). Also for the sum of PCB > 1230 ng/g lipid adjusted OR increased to 1.8 (95% CI 0.4-7.3) whereas the results were similar for DDE.