Objectives: We determined the knowledge level of water plant operators who fluoridate drinking water, and we compared small and large water plants. Methods: A pretested survey was sent to 2,381 water plant operators in 12 states that adjust the fluoride concentration of drinking water. A z-test for proportion was used to test for statistical difference between small and large plants at α=0.05. Small water plants were those treating less than 1 million gallons of water daily. Results: Eight hundred small and 480 large water plant operators responded, resulting in a response rate of 54 percent. Two-thirds of water plant operators correctly identified the optimal fluoride level, but more than 20 percent used a poor source for choosing the optimal level. Only one-fourth of operators were able to maintain the fluoride concentration to within 0.1 mg/L of the optimal concentration. A significantly greater proportion of operators at large water plants than at small water plants reported that they were able to maintain a fluoride concentration to within 0.1 mg/L of the optimal concentration (33.5% vs 21.3%, z=4.74, P <.05). Conclusions: Although most operators correctly identified the optimal fluoride level, small water plant operators were less likely to use accurate reasoning for choosing that level and in maintaining fluoride concentrations within 0.1 mg/L of that level than large water plant operators.