Despite ample pharmacological evidence for the existence of multiple mu-, delta- and kappa-opioid receptor subtypes, only three genes encoding mu-(MOR), delta-(DOR) and kappa-(KOR) opioid receptor have been cloned. The KOR gene encodes kappa(1)-sites, which specifically bind arylacetamide compounds, and the possible existence of kappa-opioid receptor subtypes derived from another kappa-opioid-receptor gene, yet to be characterized, remains a very contentious issue. kappa(2)-Opioid receptors are described as binding sites typically labelled by the non-selective benzomorphan ligand [3H]bremazocine in the presence of mu-, delta- and kappa(1)-opioid receptor blocking ligands. To investigate the genetic origin of kappa(2)-opioid receptors, we have carried out homogenate binding experiments with [3H]bremazocine in brains of single MOR-, DOR-, KOR- and double MOR/DOR-deficient mice. Scatchard analysis showed that 68+/-12% of the binding sites arise from the MOR gene, 27+/-1% from the DOR gene and 14.5+/-0.2% from the KOR gene, indicating that the three known genes account for total [3H]bremazocine binding. Experiments in the presence of mu-, delta- and kappa(1)-opioid receptor suppressor ligands further showed that non-kappa(1)-opioid receptor labelling can be accounted for by binding to both the mu- and delta-opioid receptors. Finally, [3H]bremazocine binding experiments performed on brain membranes from the triple MOR/DOR/KOR-deficient mice revealed a complete absence of binding sites, confirming definitively that no additional gene is required to explain the total population of [3H]bremazocine binding sites. Altogether the data show that the putative kappa(2)-opioid receptors are in fact a mixed population of KOR, DOR and predominantly MOR gene products.