Between May 1981 and January 1986, 130 consecutive patients with advanced untreated head and neck squamous cell cancer (HNSCC) received neoadjuvant chemotherapy (NAC) with two or three courses of the CABO combination (methotrexate 40 mg/m2 intravenously (IV) on days 1 and 15; CDDP 50 mg/m2 IV on day 4; bleomycin 10 mg IV on days 1, 8, and 15; and vincristine 2 mg IV on days 1, 8, and 15 every 3 weeks) prior to surgery and/or radiotherapy. Of the 123 patients evaluable for response to chemotherapy, 19 (15.4%) had a complete remission and 59 (48%) a partial response, yielding a 63.4% overall response rate. Response rate was significantly correlated with the performance status (PS) (P = 0.001), the stage (P = 0.005), and the T class (P = 0.02); 107 patients completed subsequent local treatment (87 with radiotherapy and 20 with surgery +/- radiotherapy). The median survival of the 124 patients evaluable for survival was 14.7 months and the overall survival rate at 3 years was 24%. The median survival and the overall survival at 3 years for the surgical subgroup were 24.7 months and 38% and for the radiotherapy subgroup were 14.3 months and 22%. These results were compared with those obtained in a historical control group of 79 patients treated in our institute with short courses of chemotherapy regimens, which did not include cisplatin, followed by radiotherapy (29 patients) or local treatments alone (26 patients with radiotherapy and 24 patients with surgery +/- radiotherapy) between January 1976 and December 1980. Most of the patient characteristics were evenly distributed (age, sex, and primary sites) except that more advanced lesions were included in the NAC group (Stage IV: 85% versus 62%; T4: 65% versus 42%; N2-3: 48% versus 29%). The overall survival was significantly higher in patients receiving NAC than in the historical control group, comparing both the groups taken as a whole (P less than 0.05) and the surgery +/- radiotherapy (P less than 0.05) and the radiotherapy (P less than 0.02) subgroups. This experience suggests a positive role of NAC on survival. However, this improvement in outcome compared with a historical control group cannot be regarded as definitive evidence for benefit to patients. Randomized studies using active regimens are required to confirm these data.