RESUMEN
We use data sets from the Cyrtodactylus brevipalmatus group with limited genetic and morphological sampling to demonstrate that not accounting for sampling error may adversely influence decisions regarding species delimitation and diagnosis. Lack of geographic sampling between the endpoints of a species range may recover notable interpopulational genetic differentiation consistent with species-level differentiation. Additionally, small population sample sizes may fail recover statistically different diagnostic morphological differences. Combined, these types of sampling error can produce results seemingly consistent with the recognition of cryptic speciesgenetically delimited populations lacking diagnostic morphological characters. This is the current situation within some lineages of the C. brevipalmatus group whereas in others, sampling error is less problematic and does not jeopardize their taxonomy. We note the potential negative effects for comparative biology as a whole if sampling error is not taken into consideration prior to constructing taxonomies.