BACKGROUND Perioperative antibiotic prophylaxis has significantly reduced wound infection rates in clean-contaminated head and neck surgical procedures but controversy still remains regarding the optimal antibiotic regime. OBJECTIVE To examine the efficacy of different antibiotics in head and neck oncological surgery prophylaxis. METHODS In this prospective, double-blind clinical trial, 189 patients with carcinoma of the upper aerodigestive tract were randomized to receive amoxicillin-clavulanate or cefazolin intravenously up to 1h before surgery and at 8-h intervals for an additional three doses. RESULTS An overall wound infection rate of 22% was observed. The infection rate in patients receiving cefazolin was 24% (22/92) vs. 21% (20/97) in those receiving amoxicillin-clavulanate; the difference was not statistically significant. Postoperative overall non-wound infection developed in 12% (22/189) patients; the rate of infection was 9.8% (9/92) in patients receiving cefazolin vs. 13.4% (13/97) in those receiving amoxicillin-clavulanate, without a statistically significant difference between the two groups. Gram-negative bacteria were more often isolated with Pseudomonas aeruginosa as the dominant species. The risk of postoperative infection was more influenced by the type of surgical procedure than by disease stage. CONCLUSIONS In clean-contaminated head and neck oncologic surgery amoxicillin-clavulanate prophylaxis was at least as efficient as cefazolin. However, when taking into account the fact that beta-lactamase containing strains have recently been spreading, amoxicillin-clavulanate should be the logical first choice.