In previous reports we presented a series of programs to calculate density estimates for univariate data (Salgado-Ugarte et al. 1993; 1995a). We also introduced a series of practical bandwidth rules for histograms, frequency polygons, and kernel density estimators, including cross-validation techniques (Salgado-Ugarte et al. 1995b). In this article we present an implementation of the smoothed bootstrap procedure of Silverman (1981) for multimodality assessment.