Despite the fact that there are many works treating new analytical approaches to monitor the transesterification reaction to produce biodiesel, only a few of them are able to quantify the minority, but important compounds, such as monoglyceride (MG) and diglyceride (DG), directly in the reaction medium for online quantification. Overcoming the lack of analytical applications in the biodiesel production, the present work proposes a new utilization of low-field proton nuclear magnetic resonance (LF-¹H NMR) and near-infrared (NIR), applying Partial Least Squares (PLS) regression models, for the online monitoring of fatty acid methyl esters, triglycerides, diglycerides and monoglycerides. The compounds presented acceptable errors of prediction employing developed PLS models for the range of concentration, usually presented during the whole reaction time, except for the 1,2-diglyceride (1,2-DG).