SUMMARYFaeces samples enriched for salmonellas in selenite F broth incubated at 43°C were subcultured to brilliant green MacConkey agar after 6 h. Four platings were made. As a control technique, each sample was enriched in selenite F incubated at 37°C for 24 h and subcultured once to brilliant green MacConkey agar. Of the specimens plated at 6 h, out of a total of 25 positive samples, 6 required more than a single subculture to reveal the presence of salmonellas. Twenty-four positive isolations were made from the selenite broths incubated at 37°C for 24 h and subcultured once only. All faecal samples positive by enrichment were also positive by direct plating. Nine samples seeded by 6 h produced only 1-2 salmonella colonies per plate. Convalescent specimens are, therefore, unlikely to be efficiently diagnosed by early subculture even if it is combined with multiple plating.River Taff samples, after pre-enrichment in buffered peptone water, were enriched in Rappaport's medium at 37°C for 6 h. Subcultures were made to four plates of brilliant green MacConkey agar. A control subculture was made at 24 h to a single plate of brilliant green MacConkey. Although multiple subculture at 6 h improved salmonella isolation, the method was not as efficient as a single plating from the same enrichment culture at 24 h. There was no evidence that subculture at 6 h combined with multiple plating had a value either for faecal examination or for the isolation of salmonellas from sewage-polluted natural water.Samples of other natural waters submitted routinely by Environmental Health Officers were pre-enriched in buffered peptone water before subculture to Rappaport's medium. The enrichment cultures were incubated at 37°C for 24 h and duplicate platings were made to brilliant green MacConkey agar. By double plating at 24 h, a gain of 16 % in salmonella recovery was made. This compares favourably with the gain obtained by subculture from Rappaport after 24 and 48 h incubation. It also saves 24 h in examination time.