Growth hormone (GH) response and baclofen levels were measured in seven healthy adult men following a 10-mg and a 20-mg dose of oral baclofen (gamma-aminobutyric acidB agonist) to determine the preferred dose in baclofen challenge studies. Multivariate analysis of variance (ANOVA) with repeated measures revealed no differences between the doses. However, when a univariate ANOVA with repeated measures was performed for each dose, the 10-mg dose showed no significant GH response over time, whereas the 20-mg dose showed a significant GH response over time. The average delta GH (change in GH from baseline) was 7.84 ng/ml (SD = 10.17) for the 10-mg dose and 3.34 ng/ml (SD = 3.64) for the 20-mg dose. The variability in the delta GH response to the 10-mg dose was significantly greater than the response to the 20-mg dose of baclofen. This variance in GH response was not explained by the differences in serum baclofen levels. Thus, a 20-mg baclofen dose appears to be preferable to a 10 mg-dose in baclofen challenge studies.