THE experimental demonstration that nitrosamines are carcinogenic, mutagenic and teratogenic (Magee, 1969) has raised the question of their potential danger to man, and prompted studies of the circumstances that might expose him to them. Nitrosamines have been found in a limited number of foods for human consumption (International Agency for Research on Cancer, 1972), and have been shown to be produced by bacteria that occur in food (CollinsThompson et al., 1972;Fong and Chan, 1973). They may be produced in vivo from nitrites and either secondary amines (Sander, Schweinsberg and Menz, 1968;Sen, Smith and Schwinghamer, 1969;Lijinsky and Epstein, 1970;Mirvish, 1970) or tertiary amines (Lijinsky, 1974) at thepH of the stomach.The possibility of bacterial production of nitrosamines from secondary amines in the presence of nitrate or nitrite in vivo at sites with a pH closer to neutrality has been raised. Production in vitro has been shown with single strains of four nitrate-reducing members of the enterobacteriaceae, Escherichia coli, E. dispar, Proteus vuEgaris and Serratia marcescens, from secondary amines at neutral p H (Sander, 1968). Hawksworth and Hill (1971) found that 27% of 37 strains of E. coli and a proportion of strains of other faecal aerobes and anaerobes were able to produce nitrosamines from nitrate and diphenylamine at a pH above 6.5. These authors (Hill and Hawksworth, 1972) consider that the most likely site of in-vivo bacterial production of nitrosamines is the urinary tract, because the substrates, nitrate and secondary amines (principally dimethylamine, DMA), are found in urine-and bacteria with both nitrate-reducing and nitrosamine-forming enzymes may be present during infection. The in-vitro production of dimethylnitrosamine (DMN) by three strains each of P. mirabilis, P. morganii and P. rettgeri from urine, and the detection of DMN in the urine of two patients with urinary-tract infections caused by P. mirabilis, supports this suggestion (Brooks et al., 1972;Thacker and Brooks, 1974). Nitrosamines produced by bacteria during urinary-tract infections may cause cancer in distant anatomical sites (Hill, Hawksworth and Tattersall, 1973; Hawksworth and Hill, 1974).The present survey was undertaken to define which of the bacteria commonly associated with urinary-tract infection are able to produce DMN. The survey of faecal bacteria by Hawksworth and Hill (1971) appears to show that production of diphenylnitrosamine is not a constant feature of all strains of any species. Similarly, we have found wide variation among strains when we tested bacteria that may cause urinary-tract infections for their ability to produce