Background: Fluoridated (F) milk schemes are employed in six countries to reduce dental caries in children. To maximise their benefits considerable uptake is required. Measuring compliance and understanding contributing factors is important in evaluating the effectiveness of schemes since it can be unclear whether reported sub-optimal fluoride (F) intakes, measured through urinary F excretion, are due to sub-optimal F contents of milks or lack of compliance with consumption. Objectives: To determine compliance with milk consumption for children receiving non-F or F milk (containing 0.5 or 0.9mgF per 189ml carton) and rationalise the use of compliance data for clinical observational or intervention studies involving F milk schemes. Research design: Partially randomised, partial cross-over study. Participants: 50 children aged 3-4 and 6-7y consuming non-F (n=50) and F milk (0.5mgF; n=15 children; 0.9mg F; n=16 children) at school. Results: Mean compliance for both non-F and F milk was >90% in each of the groups studied and showed no statistically significant difference for children using both milks. The 95% central range of proportions of milk consumed for groups of individuals was wider for 0.9mgF milk (25% to 100%) than for 0.5mgF milk (81% to 100%) although the greatest range of variation in compliance for within individual observations was seen for non-F milk consumption and in older children. Conclusion: Assessment of compliance with consumption should be included when dental efficacy of F milk consumption is being investigated or evaluated to quantify F exposure from milk. This is important, particularly if a change in the F dose of F milk might be under consideration.