Assessing the potential of nicotinic agonists as therapeutic agents has frequently relied upon single component EC(50) values obtained from studies of nicotinic receptors expressed in Xenopus oocytes. We have evaluated the validity of this approach using voltage clamp techniques. In general, agonist concentration-response plots for the alpha3beta2, alpha3beta4, alpha4-1beta2, alpha4-1beta4 and alpha7 combinations were poorly fitted by a single component Hill-equation. Improved fits were obtained with the sum of two components, although only in the case of alpha3beta4 and alpha4-1beta2 was the improvement significant regardless of the weighting method used. For the acetylcholine (ACh) concentration-response relationships of the alpha4-1beta2 combination, the two EC(50) values were 0.3 and 58.3 microM. For the alpha3beta4 combination, the two EC(50) components were 39 and 2919 microM. The 39 microM component of alpha3beta4 represented 36% of the sum of the maximum responses of both curves. This shows that for some combinations, the secondary components represent a well-separated, major population of receptors. Therefore, published EC(50) values which assume that only a single subtype of functional receptor is present may not accurately describe agonist action may therefore need to be re-evaluated.