The objective of this review is to describe and discuss the content validity of a sample of caries detection criteria reported in the literature between January 1, 1966, and May 1, 2000. Using filters to locate randomized or controlled clinical trials on dental caries, fluorides, sealants, and "restorative" care, I identified a total of 171 documents from MEDLINE and the Cochrane Collaboration's Oral Health Group (CC-OHG) special register. These articles met the following inclusion criteria: (1) Data had been collected from samples of patients or populations; and (2) dental caries was assessed clinically, and criteria were either published or described in the paper. From the selected articles, evidence tables were prepared describing each caries detection criterion. Analysis of the content validity of the criteria systems was based on evaluation of the disease process, exclusion of non-caries lesions, subjectivity, use of explorers, and drying of teeth prior to examination. This review included 29 unique criteria systems. Of those, 13 originated from the UK, 3 from the USA, 4 from Denmark, and others from the World Health Organization (WHO), Sweden, Switzerland, Norway, Netherlands, and Canada. Thirteen of the criteria systems either measured active and inactive early and cavitated lesions or defined separate criteria for smooth and occlusal tooth surfaces. Nine systems measured early as well as cavitated stages of the caries process, and 7 measured cavitation only. Eleven of the criteria systems provided explicit descriptions of the disease process measured or information on how to exclude non-caries from caries lesions. The use of explorers and drying and cleaning of teeth varied widely among the criteria. The majority of the newly developed criteria systems originated from Europe. In conclusion, this review of the content validity of the 29 criteria systems found substantial variability in disease processes measured, inclusion and exclusion criteria, and examination conditions.