Enteral feeds are increasingly prescribed for hospital patients requiring nutritional support, yet there is little to suggest (Lee, 1979; Drugs and Therapeutics Bulletin, 1980; Brown, 1981) that the infection hazard of bacterial contamination of these feeds is commanding the attention that it deserves. More than ten years ago it was shown that items of food prepared in hospital kitchens, including milk feeds and pureed food, were frequently contaminated with Gram-negative bacilli such as Escherichia coli, Pseudomonas aeruginosa and Klebsiella aerogenes (Shooter et al., 1969, 1971; Cooke et al., 1970; Montgomerie et al., 1970). In a later study it was found that as many as 68 per cent of enteral feeds, which were prepared in a hospital diet kitchen and destined for intensive care patients, were contaminated with up to lo4 klebsiellae per ml (Casewell & Phillips, 1978). But bacterial counts in excess of 106 per ml have been recorded (Montgomerie et al., 1970; Bastow, Greaves & Allison, 1982). The range of organisms reported is also extending and includes Enterobacter cloacae (Casewell, Cooper & Webster, 1981), Proteus spp., Streptococcus faecalis, Staphylococcus epidermidis and, more threateningly, Bacillus cereus and Staph. aureus (Bastow et al., 1982). Even with sterile ingredients, mixing in the diet kitchen may result in counts of up to 10s per ml (Bastow et aZ., 1982). The reservoirs of Gram-negative bacilli that have been demonstrated in diet kitchens include mixers, homogenizers, dish cloths, work surfaces, metal sieves, propylene jugs and a detergent dispenser (Montgomerie et al., 1970; Casewell & Phillips, 1978; Casewell et al., 1981). The elimination of all these reservoirs is probably impossible. Inoculated feeds may not reach the patient until an overnight delay, and the change from bolus to continuous-drip nasogastric feeding over a further 12-24 h provides ample opportunity for multiplication of organisms in this very adequate culture medium (Simmons, 1981). What, then, are the theoretical and actual consequences of ingesting these contaminated feeds? Firstly, there seems to be no reason why, sooner or later, enteric pathogens will not be involved with obvious consequences. Indeed, hospital-prepared feeds have already been suspected, although not proven, as the cause of hospital-acquired gastroenteritis caused by Salmonella enteritidis of obscure origin (Gill & Gill, 1981), and Staph. aweus and B. cereus have also been found in feeds (Bastow et al., 1982). Secondly, even in the absence of contamination with more aggressive pathogens, it is clear that the number of organisms ingested often exceeds the number,