Intra-arterial neoadjuvant chemotherapy (TPP) with pirarubicin, cisplatin and peplomycin produced strong primary effects on oral squamous cell carcinoma, using metoclopramide (MCA) as an anti-emetic. After clinical application of granisetron (GRN), the clinical responses to TPP observed previously were weakened. In this paper, the influence of GRN on TPP is discussed. Sixty-three cases were evaluated with regard to the primary effects of TPP and anti-emetics. GRN was used in 42 cases of the GRN group, and MCA in 21 cases of non-GRN group. The clinical response rate (complete response, CR or partial response, PR) was 95.2% in the non-GRN group, and 76.2% in the GRN group. The rate of CR in the non-GRN group was 47.6%, whereas it was 9.5% in the GRN group. The histological effects in the GRN group were significantly lower (P<0.05) than those of non-GRN group. Concerning the relationship between the clinical responses and the histological responses, 4 of the 18 CR+PR cases (22.2%) in the GRN group showed good histological responses, compared with 6 of the 14 CR+PR cases (42.9%) showing in the non-GRN group. The histological responses in the GRN group were significantly lower (P<0.05) than in the non-GRN group. Our data indicate that GRN reduces the clinical and histological responses of chemotherapy.