Groups of 33 chickens were fed continuously on diets containing feed additives that are employed commercially for a variety of purposes, and were infected orally when 4 days old with a nalidixic acid-resistant mutant of Salmonella typhimurium. The amount of S. typhimurium organisms excreted in their faeces was estimated by culturing them at intervals and in a standard manner on brilliant green agar containing sodium nalidixate; when the chickens were killed their caecal contents were examined by the same technique.

Avoparcin and lincomycin, like nitrovin and tylosin (Smith & Tucker, 1975b), favoured colonization of the alimentary tract by the S. typhimurium organisms as shown by the fact that the chickens to which they were fed excreted these organisms in their faeces in higher concentration and for longer periods of time than did chickens fed on non-medicated diets. Amprolium, monensin, dimetridazole, arsenilic acid and nitro-hydroxyphenylarsonate had no obvious effect on the salmonella excretion pattern.

When only five chickens in each group were experimentally infected so that the effect of the feed additives on infections acquired by contact could be monitored, avoparcin, lincomycin, nitrovin and tylosin again favoured colonization of the alimentary tract with the S. typhimurium organisms and so did dimetridazole. Arsenilic acid, in contrast, hindered the development of infection. Amprolium, monensin and nitro-hydroxyphenylarsonate were without obvious effect.

Many of the chickens that were fed on diets that favoured S. typhimurium colonization, but not those fed on non-medicated diets, were still excreting S. typhimurium organisms in their faeces when they were killed at 56 days of age, the age at which broiler chickens kept under commercial conditions are usually slaughtered.