Data are presented on a study carried out on water analyzed with three different methods. Thus, a comparison was performed in parallel with the ISO 9308-1 reference method, stated by the European Drinking Water Directive, and two enzyme-based methods, the Colilert 18/Quanti-Tray (TM) and the Chromogenic E. coli Agar. Phenotypical characteristics of presumptive E. coli and confirmation tests of the isolates were evaluated. The selectivity of the ISO method, combined with selective confirmation steps, has resulted in lower E. coli counts in comparison with the two alternative methods. In fact, the method failed to detect a significant proportion of the target microorganisms; furthermore, it allowed the growth of other bacterial species, mainly Klebsiella oxytoca. The two alternative methods were more sensitive than the ISO procedure and were able to sustain the growth of different fractions of bacterial populations. Particularly, the Colilert was able to detect qualitatively and quantitatively a higher proportion of E. coli. Furthermore, the hydrolysis of 4-methyl-umbelliferyl-beta-D-glucuronide as a marker for E. coli seems to be a more appropriate procedure for the identification of the target microorganisms with respect to the more traditional test for indole production.