Two methods of bacteriocin susceptibility typing for Serratia marcescens were compared. A total of 80 epidemiologically unrelated isolates from patients in a single hospital were typed by the cross-streaking method and the mitomycin C-induced (spotting) method. The cross-streaking method was found to be more discriminatory than the spotting method. Using the cross-streaking method, it was possible to differentiate 50 bacteriocin groups out of the 80 isolates, whereas only 31 groups could be obtained with the spotting method. The reproducibility and percentage typability of the cross-streaking method (82.5 and 93.75%, respectively) were found to be as good as, if not better than, those of the spotting method (78.75 and 90.0%, respectively). Other factors, such as lower economic cost, technical simplicity, and the relative ease in the scoring of results, indicate a preference for the cross-streaking method. The findings of this study support the choice of the cross-streaking method for the bacteriocin typing of S. marcescens in epidemiological studies.