Recent analyses of Sepkoski's genus-level compendium show that only three events form a statistically separate class of high extinction intensities when only post-Early Ordovician intervals are considered, but geologists have called numerous events mass extinctions. Is this a conflict? A review of different methods of tabulating data from the Sepkoski database reveals 18 intervals during the Phanerozoic have peaks of both magnitude and rate of extinction that appear in each tabulating scheme. These intervals all fit Sepkoski's definition of mass extinction. However, they vary widely in timing and effect of extinction, demonstrating that mass extinctions are not a homogeneous group of events. No consensus has been reached on the kill mechanism for any marine mass extinction. In fact, adequate data on timing in ecologic, rather than geologic, time and on geographic and environmental distribution of extinction have not yet been systematically compiled for any extinction event.