Average-sized portions of a variety of food products were reacted with nitrite under realistically simulated gastric conditions. The aqueous incubation medium contained sodium nitrite (10 mg/l) and potassium thiocyanate to mimic the incoming flux of saliva, as well as pepsin, sodium chloride and hydrochloric acid, reflecting the composition of gastric juice. After incubation for 2 hr at 37 degrees C, volatile N-nitrosamines and N-nitrosamino acids were determined in the reaction mixtures. Nitrosodimethylamine (NDMA) was present in the incubation mixtures of smoked mackerel (8.5 micrograms per portion), canned herring (0.66 micrograms per portion) and beer (0.70 micrograms per 'portion'). Smaller amounts per portion, sometimes of other nitrosamines as well, were observed with canned salmon and anchovy, mustard, yoghurt and coffee brew. Negative results were obtained for canned tuna, soya sauce, ketchup, white bread, 'nasi goreng', tea brew and cocoa milk. Nitrosamino acids were detected in the reaction mixtures of smoked mackerel (58 micrograms per portion), soya sauce (24 micrograms per portion) and canned salmon (6.9 micrograms per portion) and in smaller amounts in those of canned herring, anchovy and cocoa milk. In order to reduce the number of analyses to be performed, most products have been studied only after incubation, so that the nitrosamines and nitrosamino acids found may already have been present -- wholly or partly -- in the original products, before incubation. Such is the case for part of the NDMA in the reaction mixture of smoked mackerel and for all the NDMA in beer. The toxicological implications of these findings remain to be established.