In situ (intra-oral) caries models are used for two purposes. First, they provide information about oral physiological processes. Such information helps to detail our knowledge of the oral ecosystem and to verify conclusions from in vitro experiments. Second, in situ models are utilized to test preventive agents in the phase between laboratory testing and clinical trials. Most investigations involving enamel inserts have been aimed at testing new dentifrices. The experimental designs of such studies usually do not allow one to draw conclusions on physico-chemical processes, e.g., because of single point measurements. Studies of model parameters (lesion type, lesion severity, and de/remineralization in time) constitute only a minority of the research reports. The most striking observation obtained with in situ models has been the significant differences in de/remineralization observed among individuals and, more importantly, within one individual during different time periods and between different sites in the same mouth (for review, see ten Cate et al., 1992). Regardless of this, some general findings can be inferred: During in situ demineralization, up to 62 vol% microns/day may be removed from enamel. For dentin specimens, this value may be as high as 89 vol% microns/day. For remineralization, during fluoride dentifrice treatment, a median deposition rate of 0.7%/day (for lesions with integrated mineral loss values between 2000 and 4000 vol% microns) is found. The rate of deposition seems to be correlated with the extent of the pre-formed lesion. This suggests that the number of sites (crystallite surface) available for calcium phosphate precipitation is an important parameter.(ABSTRACT TRUNCATED AT 250 WORDS)