Two experiments were carried out to investigate possible causes of nicarbazin residues in broiler chicken tissues. The first experiment was designed to establish whether feeding nicarbazin as stipulated in the product license can result in 4,4'-dinitrocarbanilide (DNC) tissue residues exceeding the JECFA MRL (200 micrograms/kg). It was shown that the MRL was exceeded in the livers of broilers housed on deep litter, but not in those of broilers housed on wire flooring. Muscle DNC concentrations were well below the MRL. The higher residual tissue concentrations in birds housed on deep litter were attributed to faecal recycling. The second experiment was to establish the relationship between nicarbazin-contaminated withdrawal ration up to the point of slaughter and DNC residues in the tissues of broilers that had not been previously exposed to nicarbazin. Tissue DNC concentrations were found to be proportional to feed concentrations. The housing method caused no significant difference in tissue residues. Meal containing nicarbazin at a concentration of 2.4 mg/kg or greater caused liver DNC residues above the JECFA MRL. Violative residues may, therefore, occur in chickens not exposed to nicarbazin during rearing, but fed withdrawal ration contaminated at 2.4 mg/kg or greater, or in chickens housed on deep litter and fed nicarbazin-medicated meal according to the product license even when the withdrawal ration is nicarbazin-free.