Utah's Height and Weight Measurement Project was conducted with elementary school students periodically from 2002 to 2008. The 2002 pilot project was performed to establish variability rates between schools and within schools. It allowed us to accurately determine both the sample size and the number of schools that were required to calculate a reliable statewide estimate based on a random sample of schools and to establish sentinel grades. The sentinel grades identified were grades 1, 3, and 5. Use of randomly selected classes in the sentinel grades resulted in decreased sample size and less school disruption while maintaining sufficient precision. Standardized, calibrated equipment was purchased and given to school nurses for safekeeping. Lessons learned included establishing strong relationships with partners, especially school nurses, and obtaining support from upper management at the schools, school districts, and the Utah Department of Health. A significant difference in participation rates and obesity rates at the individual school level was observed depending on parental consent type; active consent was associated with lower student participation rates and lower observed obesity rates. Data were presented to both participating and nonparticipating schools, school nurses, district superintendents, and principals. For surveillance purposes, sampling is an efficient, cost-effective way to estimate childhood overweight and obesity rates.