SortPred is a random forest based easy to use standalone program for the prediction of bacterial sortases. SortPred was developed by using different feature sets that includes, amino acid composition, dipeptide composition, physiochemical properties, and their hybrids as an input feature. The SortPred working module consists of two layers. Firstly, Layer 1 of SortPred predicts whether a given sequence is a sortase or non-sortase. In the second step, if the given sequence is predicted to be a sortase, Layer 2 predicts its class (A, B, C, D, E or F). In both the cases the probability scores are also provided.