In previous studies, Bacillus subtilis has been used to control mould growth during red sorghum malting. The use of this biocontrol in steeping liquor has been optimized with some success and the combined use of a 0.2% NaOH steep, followed by a re-steep in a B. subtilis-based biocontrol, has been proposed. The sharpness and variability of the β-amylase peak and the higher levels of β-glucanase, obtained in the presence of B. subtilis cells, were highlighted. In this work, the suitability of the Weibull four-parameters model to predict sorghum malt α-amylase activity during the enzyme induction stage of red sorghum germination has been compared with that of a second order polynomial model and a general linear model. Results obtained showed that the Weibull four-parameters model could be used to predict α-amylase activity, with significant goodness of fit when compared with the second order polynomial model and the general linear model. The effects of steeping treatment (combined use of 0.2% NaOH and B. subtilis-S499 starters) and the germination temperature are presented. When the B. subtilis culture used as a starter was diluted, the treatment efficacy to develop α-amylase activity was lost. This study also showed that the germination temperature affected the α-amylase activity rate increase during the induction phase.