Current theories on the causes of extinction at the Cretaceous-Tertiary boundary have been based on previously published data; however, few workers have stopped to ask the question, 'How good is the basic data set?' To test the accuracy of the published record, a quantitative and qualitative analysis of the Crassatellidae (Mollusca, Bivalvia) of the Gulf and Mid-Atlantic Coastal Plains of the United States for the Upper Cretaceous and lower Tertiary was conducted. Thirty-eight species names and four generic names are used in publications for the Crassatellidae within the geographic and stratigraphic constraints of this analysis. Fourteen of the 38 species names are represented by statistically valid numbers of specimens and were tested by using canonical discriminant analysis. All 38 names, with the exception of 1 invalid name and 4 names for which no representative specimen could be located, were evaluated qualitatively. The results show that the published fossil record is highly inaccurate. Only 8 valid, recognizable species exist in the Crassatellidae within the limits of this study, 14 names are synonymized, and 11 names are represented by indeterminate molds or poorly preserved specimens. Three of the four genera are well founded; the fourth is based on the juvenile of another genus and therefore synonymized. This detailed taxonomic analysis of the Crassatellidae illustrates that the published fossil record is not reliable. Calculations of evolutionary and paleobiologic significance based on poorly defined, overly split fossil groups, such as the Crassatellidae, are biased in the following ways: Rates of evolution and extinction are higher, Faunal turnover at mass extinctions appears more catastrophic, Species diversity is high, Average species durations are shortened, and Geographic ranges are restricted. The data on the taxonomically standardized Crassatellidae show evolutionary rates one-quarter to one-half that of the published fossil record; faunal change at the Cretaceous-Tertiary boundary that was not catastrophic; a constant number of species on each side of the Cretaceous-Tertiary boundary; a decrease in abundance in the Tertiary; and lower species diversity, longer average species durations, and expanded geographic ranges. Similar detailed taxonomic studies need to be conducted on other groups of organisms to test the patterns illustrated for the Crassatellidae and to determine the extent and direction of the bias in the published fossil record. Answers to our questions about evolutionary change cannot be found in the literature but rather with the fossils themselves. Evolution and extinction occur within small populations of species groups, and it is only through detailed analysis of these groups that we can achieve an understanding of the causes and effects of evolution and extinction.