Arsenic and cadmium are important inorganic toxicants in the environment. Humans certainly have the potential to be exposed to the mixtures of arsenic and cadmium, but the toxicological interactions of these inorganic mixtures are poorly defined. A general population co-exposed to arsenic and cadmium, was selected in China. The total number of participants was 245, made up of 122 in the arsenic-cadmium polluted area, 123 in the non polluted area. Urinary arsenic (UAs) and cadmium (UCd) were determined by atomic absorption spectrometry as exposure biomarkers and beta2-microglobulin (Ubeta2MG), albumin (UALB), N-acetyl-beta2-glucosaminidase (UNAG) in urine were determined as effect biomarkers. The benchmark dose (BMD) and the lower confidence limit on the benchmark dose (LBMD) were calculated to estimate the critical concentration of UAs and UCd. UAs and UCd concentrations in the polluted area were significantly higher than those in the non polluted area (P < 0.01). The levels of Ubeta2MG, UALB and UNAG in the polluted area were significantly higher than those in the non polluted area (P < 0.01). The BMD/LBMD of UAs and UCd for a 10% level of risk above the background level were estimated as 121.91/102.11 microg/g creatinine and 1.05/0.88 microg/g creatinine. It was suggested that the lower confidence limit of population critical concentration of UAs and UCd for renal dysfunction for 10% excess risk level above the background, which is obtained from LBMD, may need to be kept below 102 and 0.88 microg/g creatinine in order to prevent renal damage in general population co-exposed to arsenic and cadmium. It is indicated that combined effect of arsenic and cadmium were additive effect and/or synergistic effect, and cadmium may potentiate arsenic nephrotoxicity during the long-term and co-exposure to arsenic and cadmium in humans.