Genetic differentiation among populations of several organisms has been explored using AFLP. Cluster analysis methods have been applied to visualize differences in AFLP and to test hypotheses in ecology and evolution. However, this approach is potentially flawed because it is unknown whether the number of individuals and markers used are sufficient to validly represent the genome of subject populations. We present a method for determining the number of individuals and the number of markers to consider when representing genetic profiles. The proposed method uses bootstrapping to construct curves representing the standard error of similarity matrices (SESim curves) at different number of individuals by number of marker combinations. The point at which a SESim curve reach a value smaller than 0.045 indicates the minimum number of individuals by markers to use. This method is used in our genotypic differentiation study of a braconid. AFLP data from the processionary moth, was used to validate the method.