The modifying effects of three doses of dietary protocatechuic acid (PCA) given the initiation and postinitiation phases of oral carcinogenesis initiated with 4-nitroquinoline 1-oxide (4-NQO) were investigated in male F344 rats. At 6 weeks of age, rats were divided into experimental and control groups and fed the diet containing PCA at various doses of 0 g/kg diet (basal diet alone), 0.5 g/kg diet (500 ppm), 1 g/kg diet (1000 ppm), and 2 g/kg diet (2000 ppm). At 7 weeks of age, all animals except PCA alone and control groups were given 4-NQO (20 ppm) in the drinking water for 8 weeks to induce oral cancer. Seven days after the 4-NQO exposure, groups of animals fed the PCA diets were switched to the basal diet and continued on this diet until the end of the study. Starting 1 week after the end of 4-NQO exposure, the groups given 4-NQO and a basal diet were switched to the diets containing PCA and maintained on these diets for 22 weeks. The other groups consisted of rats given 2000 ppm PCA alone or untreated rats. All animals were necropsied at the termination of the experiment (week 32). The incidences of tongue neoplasms and preneoplastic lesions, polyamine levels in the tongue tissue, and cell proliferation activity estimated by bromodeoxyuridine-labeling index and by morphometric analysis of silver-stained nucleolar organizer regions' protein were compared among the groups. Feeding of PCA at all doses during initiation or postinitiation phase significantly decreased the development of tongue neoplasms (squamous cell papilloma and carcinoma) and preneoplasia (hyperplasia and dysplasia) (P < 0.05). There were no such lesions in rats treated with 2000 ppm PCA alone or those in an untreated control group. Dietary administration of PCA also caused significant decreases in the labeling index of bromodeoxyuridine and the number and area of silver-stained nucleolar organizer regions per cell nucleus, known as cell proliferation indices, of the tongue squamous epithelium (P < 0.05). In addition, PCA exposure during either initiation or postinitiation phase decreased polyamine levels in the oral mucosa (P < 0.05). These results clearly indicated that PCA inhibited rat oral carcinogenesis in both initiation and postinitiation phases, when administered in these respective phases together with or following treatment with 4-NQO, and such inhibition might be related to suppression of cell proliferation by PCA.