This study addresses the glyceride transesterification (TE) kinetics, a crucial step in biodiesel production from vegetable oils. An nth order reversible model is considered to describe the TE reaction rate, using experimental data gathered from different authors for a broad range of conditions. An incremental model-building strategy is used, consisting of the following sequence of steps: (a) structural identifiability analysis; (b) outliers detection using a robust M-estimator; (c) parameter estimation and subsequent construction of the respective confidence intervals; and (d) practical identifiability analysis. The methodology is applied to the glyceride TE reaction with sodium methoxide as catalyst, using a data set comprising 144 points collected from literature, and comparing the results for a five (5P) and a six (6P) parameters' model. Although the results show that both kinetic models are structurally nonidentifiable, the former is practically identifiable for the case study analyzed. Furthermore, the 5P model shows a good capability of explaining the data features, with an average relative deviation of 5.2% for the residuals. The methodology presented can be used to analyze general kinetic reactions when experimental data are available and plausible reaction mechanisms are known.