Special diets that were extremely low in protein (5.5%) and high in carbohydrate were used to test the effect of nicotinamide on N-nitrosodimethylamine-induced carcinogenesis in Holtzman albino rats. The level of nicotinamide in the three diets ranged from 0 mg/kg of food to 50 mg/kg to 500 mg/kg. During the treatment with these diets (5 weeks) and the carcinogen (18 days), the renal and hepatic concentrations of nicotinamide adenine dinucleotide-reduced nicotinamide adenine dinucleotide and nicotinamide adenine dinucleotide phosphate-reduced nicotinamide adenine dinucleotide phosphate were measured. In the liver, the concentration of these coenzymes fell well below normal levels, and significant differences between the two extreme diets (0 and 500 mg/kg) were found in the hepatic content of nicotinamide adenine dinucleotide-reduced nicotinamide adenine dinucleotide. In the kidney, the treatment with special diets and carcinogen had less effect. The content of nicotinamide adenine dinucleotide phosphate-reduced nicotinamide adenine dinucleotide phosphate remained normal, and the only significant drop in the concentrations of nicotinamide adenine dinucleotide-reduced nicotinamide adenine dinucleotide phosphate occurred in the animals on the nicotinamide-deficient diet. After the treatment, all of the animals were returned to the same standard diet. During the remainder of the experiment, 85 weeks, it was found that the initial treatments had not affected tumor incidence levels or tumor type, but had altered the rate of tumor incidence. These differences could be seen by comparing the results for the rats that had been given the nicotinamide-deficient diet to the results for the animals receiving an excess of the vitamin.