The aim was to assess caries increment as influenced by partial substitution of sucrose by xylitol (X group) over a 2-year period in comparison with systemic fluoride (F group) and restorative treatment only (C group). The study differed from the 3-year field study of the same series primarily in that existing base-line differences were eliminated because the protocol required that all the new subjects entering the institutions in the 1st year were to be included for a 2-year trial. During this period the number of dropouts was 243 (19.9% of all subjects), the final material consisting of 976 children (6-12 years old). The 2-year DMFS increment was 3.8 in the X group, 4.8 in the F group, and 6.0 in the C group. The corresponding ratio (RS) between caries incidence and the tooth surface population at risk was RSX, 4.5; RSF, 5.5; and RSC, 7.5. The xylitol regimen resulted in a lower increment of caries than measured in the F and C groups (p less than 0.001; convariance analysis, with base-line prevalence, number of permanent teeth, and visible plaque index as covariants.