In order to determine the suitability of a large group of commercial surfactants for applications where small amounts may enter food, the toxicity of 22 products was studied. Ethylene oxide chain lengths ranged from 4 to 40, and alkylphenols included di(secondary butyl), octyl, nonyl, and dodecyl. Three products were administered to rats and dogs in the diet for 2 years, 13 to rats for 90 days, and 8 to dogs for 90 days. Absorption and elimination of 2 products was studied in rats. No effect was found as a result of 90-day feeding of: nonyl 4-rats and dogs 0.04 g/kg/day; nonyl 6-rats <0.04, dogs 0.04; dodecyl 6-<0.04; octyl 9-rats 0.2, dogs <0.04; nonyl 9-rats 0.01, dogs 0.04% in diet; dodecyl 9-rats 0.04; nonyl 15-rats and dogs 0.04; nonyl 20-rats 1.0, dogs <0.04; nonyl 30-rats 5.0, dogs 1.0; octyl 40-rats and dogs 5% in diet; nonyl 40-rats 0.3% in diet; di(sec-butyl) 40-rats 1% in diet; dodecyl 40-rats 5.0. The effect at the next highest level fed was focal myocardial necrosis for nonyl 20, but only trivial for the other samples; softness of intestinal contents, low food intake, emesis, somewhat elevated liver to body weight ratio. No effect was found in 2 years from nonylphenol ethylene oxide 9 on rats at 0.14 or dogs at 0.03 g/kg/day in the diet; from nonylphenol ethylene oxide 4 at 0.2 and 0.04, respectively; or from octylphenol ethylene 40 on rats at 0.7 g/kg/day. No carcinogenic effect was found in any feeding test. In dogs but not in rats there was nonprogressive myocardial toxicity from oral doses, apparently a direct pharmacodynamic effect on the heart muscle, proportional to the content of ethers with ethylene oxide chains 20 units in length. This was marked in products with an average chain length of 20, but not detected with averages of 12 or 25. © 1969.