1. For several years, we have used the cholecystokinin (CCK)B/gastrin receptor selective antagonist, L-365,260, as a reference compound in a variety of studies in CCKB/gastrin receptor radioligand binding assays. Here, we have analysed the competition curve data sets obtained between L-365,260 and [125I]-BH-CCK8S in guinea-pig gastric gland and mouse and rat cerebral cortex preparations. 2. Competition curves obtained for L-365,260 in the mouse cortex assay were not different from rectangular hyperbolae (slope = 1.01 +/- 0.02) implying the presence of a single population of binding sites (pKI = 8.41 +/- 0.01; data from 47 experiments, slope constrained to unity). However, in the rat cortex and guinea-pig gastric gland assays, the mean slope of the competition curves was significantly less than one and the mean apparent pKI significantly lower than that obtained in the mouse cortex (slope = 0.85 +/- 0.03, 0.90 +/- 0.03; apparent pKI = 7.98 +/- 0.05, 8.07 +/- 0.05; 48 and 45 experiments, in rat and guinea-pig, respectively). The distribution of the individual pKI and slope estimates of the competition curves in these two assays was consistent with expectations for the variable expression (in terms of absolute number and proportion) of two binding sites. The two sites were characterized by pKI values for L-365,260 of 8.50 +/- 0.04 and 8.48 +/- 0.04 for the high affinity site and 7.32 +/- 0.04 and 7.22 +/- 0.06 for the low affinity site in guinea-pig and rat, respectively. 3. The affinity estimates for L-365,260, although obtained on different tissues, are consistent with data obtained from the analysis of L-365,260 antagonism of pentagastrin-stimulated responses in mouse and rat stomach (acid secretion) and guinea-pig gastric muscle (isotonic contraction) assays. To this extent, these data suggest the existence of two CCKB/gastrin receptor subtypes.