The aim of this study was, systematically, to evaluate the effect of dietary changes in the prevention of dental caries. A search and analysis strategy was followed, as suggested by the Swedish Council on Technology Assessment in Health Care (SBU). The search strategy for articles published in 1966-2003 was performed using electronic databases and reference lists of articles and selected textbooks. Out of 714 articles originally identified, 18 met the inclusion criteria for a randomized or controlled clinical trial--at least 2 years' follow-up and caries increment as a primary endpoint. This included the total or partial substitution of sucrose with sugar substitutes or the addition of protective foods to chewing gum. No study was found evaluating the effect of information designed to reduce sugar intake/frequency as a single preventive measure. It is suggested that the evidence for the use of sorbitol or xylitol in chewing gum, or for the use of invert sugar, is inconclusive. No caries-preventive effect was found from adding calcium phosphate or dicalcium phosphate dihydrate to chewing gums. The review dearly demonstrates the need for well-designed randomized clinical studies with adequate control groups and high compliance.