Three methods for detecting E. coli in marine waters with the Multiple Tube Dilution method with MUG supplement: (1) Lauryl Tryptone Broth (LTB) at 35°C then EC-MUG broth at 44.5°C; (2) LTB-MUG broth at 35°C then EC broth at 44.5°C; (3) EC-MUG broth at 35°C then EC broth at 44.5°C, incubated to 48 h, were compared. E. coli was confirmed using Eosin Methylene Blue agar and IMViC reactions or API20E. Method (1) detected more E. coli positive tubes (158) than methods (2) (135) or (3) (142). Six of the 158 tubes were fluorescent negative but E. coli positive (3.8% false negative) and eight were fluorescent positive but E. coli negative (5.1% false positive). Incubation for 24 h gave 5.7% false negatives and 3.8% false positives, whereas incubation for 48 h gave 3.8% false negatives and 5.1% false positives. In (1) gas production gave 24.7% false positives and 13.3% false negatives. Of 131 E. coli isolates retested for gas production and MUG reaction, the MUG reaction was best in EC-MUG at 35°C (5.3% weak fluorescence) and worst in LTB-MUG (23.7% weak fluorescence). For E. coli in marine waters the MUG reaction without confirmation gave satisfactory results, and gas production had no value in their presumptive detection.