Purpose: Studies of the effects of perioperative dexamethasone (DEX) during oncologic surgery are scarce. The first aim of the present study was to clarify whether perioperative DEX affects the shortterm mortality in patients with head and neck cancer (HNC). The second aim was to analyze the causes of death and predictors affecting long-term mortality. Patients and Methods: The present prospective, double-blind randomized, controlled study included patients with HNC who had undergone microvascular reconstruction from 2008 through 2013. The patients were randomized into 2 groups: the receipt of perioperative DEX for 3 days (study group) or no DEX (control group). The patients' data and cause of death were registered until the end of 2017. The primary cause of death was used in the analyses. Results: A total of 93 patients were included in the present study: 51 in the DEX group (study group) and 42 in the NON-DEX group (control group). Altogether 38 patients died during a median follow-up period of 5.3 years. During the first year, more deaths had occurred in the DEX group than in the NON-DEX group: at 1 month, 4% versus 0%; at 6 months, 14% versus 0%; and at 12 months, 22% versus 5% (P = .043). The overall survival rate for all patients was 59%. HNC was the primary cause of death for most of the patients who died. On univariate analysis, the deceased patients had more advanced disease (higher T classification, P = .002; higher stage, P = .008), a greater need for a gastrostoma (P = .002), more often received postoperative chemotherapy (P = .005), and more often had locoregional (P = .025) or distal (P < .001) metastases. In the multivariate Cox model, the most important long-term predictors of death were the presence of