The objective of this study was to compare the gain-response curve to dietary levels of 2-hydroxy-4(methylthio) butanoic acid (HMTBA) and DL-Met (DLM) across 4 floor pen trials in which different diets were used. Six replicates of 38 or 41 birds per pen (trials 1 to 2 and 3 to 4, respectively) were used in a 2 x 3 factorial arrangement. A control with 12 replicates was also included. The 2 Met sources were fed at 3 equimolar levels equally spaced, with the highest level added at requirements from 1 to 48, 49, 43, or 49 d for trials 1, 2, 3, and 4, respectively. Commercial-type TSAA-deficient control diets contained sorghum, wheat, corn, or corn plus meat and bone meal for trials 1, 2, 3, and 4, respectively. Performance improved at all times for most parameters after supplementing with HMTBA or DLM (P < 0.05). No differences were found in the birds fed HMTBA or DLM at any age and trial (P > 0.05), except for trial 1, in which 17-d-old birds performed better when fed HMTBA than DLM (P < 0.05). In each trial, linear, quadratic, and exponential regressions were conducted upon the gain response of birds fed HMTBA and DLM separately. Equations with better goodness of fit were used to compare the estimated gain responses to feeding HMTBA vs. DLM. In 3 trials, the shape of the gain-response curve differed when feeding HMTBA vs. DLM. In trials 3 and 4, feeding HMTBA at commercial levels resulted in greater gain responses than DLM (P < 0.05), whereas, in trials 2 and 4, at very deficient levels, DLM-fed birds outperformed those fed HMTBA (P < 0.05). When the 4 trials were combined, the dose-response curve with the best goodness of fit was linear for HMTBA and quadratic for DLM. It can be concluded that the 2 Met sources have a different dose-response form, HMTBA could outperform DLM at commercial levels, and DLM could outperform HMTBA at deficient levels.