Various multiparticulate dissolution models that assume a log-normal particle-size distribution are fitted by nonlinear least-squares regression to data from the dissolution of micronized glyburide. Estimates of parameters describing the effective initial particle-size distribution are obtained, together with estimates of the specific dissolution rate parameter. A dissolution equation based on an ideal, untruncated, log-normal distribution, with the single particles dissolving according to the Hixson-Crowell cube root law, is the best model. The dissolution behavior of glyburide can be well described by this model in terms of the specific dissolution rate parameter and one other parameter accounting for the distribution effect. The estimation of these two parameters represents a more exact way of describing the dissolution characteristics of drug powders than previous approaches. The method should be of interest in the quality control of drugs that may cause bioavailability problems because of dissolution rate-limited absorption.