Objectives. We compared insulin-like growth factor (IGF-I) levels obtained in two groups with different methods of assessing compliance. Burned children were randomized to receive a daily injection of 0.05-0.1mg/kg per day of recombinant human growth hormone (rhGH) or placebo. Study design. One hundred twenty-five children (age range 6 months to 17 years) with total body surface area burns >40% participated in the study. Baseline levels of IGF-I were obtained at hospital discharge, at which time daily injections of rhGH or placebo were initiated. Assessment of IGF-I levels was repeated at 3-month intervals for 1 year. A directly monitored group met daily with research staff that witnessed the preparation and injection of the study drug. A self-reported group completed a Self Reported Compliance Questionnaire (SRCQ) that assessed compliance with drug regimen. A compliance of at least 85% (injections reported administered/maximal total injections possible) was accepted as being compliant. Data were analyzed using a one-way ANOVA followed by a Student Newman-Kuels test, with the results given as means+/-S.E.M. Results. The percent change in IGF-I levels between 6 and 9 months in the self-reported (22.7+/-7.9%) and monitored groups (21.9+/-8.6%) were similar. In contrast, the percent change in IGF-I levels in the placebo group was significantly decreased (-5.6+/-6.3%). Conclusions. Self-reported scores via an SRCQ is a potentially useful and valid method of assessing compliance of rhGH injections, as both reported and directly monitored methods yield similar changes in levels of IGF-I.