In the present paper a factor analysis is presented for the enzymatic transesterification of waste oil for biodiesel production. The experimental data on batch reactor evidence two key variables: enzyme loading and mixing conditions. These variables were subjected to a factor analysis and their combined effect on the reaction performance was determined. Response surface methodology (RSM) was used based on a linear first order model (steepest ascent method) and on a second order one in proximity of the optimal Solution. The result was a model able to predict reaction performance within the range of mixing rates and enzyme amount considered for model formulation and outside of it, as shown in the final validation. Best performances were obtained at high stirring and high enzyme loading. (C) 2009 Elsevier Ltd. All rights reserved.