The aims of this study were to estimate the caries-preventive effects of a school-based weekly fluoride mouthrinse (FMR) program and to determine whether its effectiveness varied by school-level caries risk. We used clinical and parent-reported data for 1,363 children in grades 1 through 5 from a probability sample of North Carolina (NC) schoolchildren. Children's caries experience was measured using decayed and filled primary (d(2,3)fs) and total (d(2,3)fs+D(2,3)MFS) tooth surfaces. Program participation was quantified using 'FMR years'. To estimate caries risk at program entry, children were matched with NC kindergarten-surveillance data representing school-level mean untreated decay (low-risk school: < 1 and high-risk school: ≥ 1 untreated carious teeth). Mean d(2,3)fs was 4.1 [95% confidence limits (CL) = 3.7, 4.5]. Overall, each 'FMR year' was associated with weak reduction of caries prevalence in the primary [prevalence ratio (PR) = 0.98; 95% CL = 0.90, 1.06] and the mixed dentition (PR = 0.98; 95% CL = 0.91, 1.05). We found a trend toward a larger caries-preventive benefit among children in high-risk schools compared with those in low-risk schools (i.e., 55% vs. 10% caries reduction for 5 to 6 yrs of FMR participation compared to none). Although this difference was not confirmed statistically, our results indicate that children in high-risk schools, as identified by school-level surveillance data, may experience substantial caries-preventive benefits from long-term FMR participation.