The purpose was to study differences in the caries increment rate as influenced by various sugars. The trial involved almost complete substitution of sucrose (S) by fructose (F) or xylitol (X) during a period of 2 years. There were no significant initial differences as to caries status between the prospective sugar groups; 35 subjects in the S-group, 38 in the F-group, and 52 in the X-group. During the entire study 10 subjects discontinued or were excluded. The clinical and radiographical observer error was reported and discussed. After 2 years the mean increment of decayed, missed and filled tooth surfaces was 7.2 in the S-group, 3.8 in the F-group, and 0.0 in the X-group. The weakness of the DMFS-index in not showing the development of new secondary caries and the increase in size of the lesions was overcome by expressing the caries activity in terms of indices showing the total quantitative and qualitative development. The results showed a massive reduction of the caries increment in relation to xylitol consumption. Fructose was found to be less cariogenic than sucrose. It was suggested that the non- and anticariogenic properties of xylitol principally depend on its lack of suitability for microbial metabolism and physico-chemical effects in plaque and saliva.