Introduction The FACE-Q Head and Neck Cancer Module is a comprehensive disease-specific patient-reported outcome measure that captures multiple health-related quality of life domains relevant to head and neck cancer patients. However, its disease-specific nature precludes its use in economic evaluations requiring generic preference-based utilities. This study developed and validated a mapping algorithm to predict EQ-5D-5L utilities from FACE-Q scores, enabling quality-adjusted life year (QALY) calculations and cost-utility analyses for patients undergoing jaw reconstruction. Methodology Data were collected from patients undergoing jaw reconstruction following cancer resection at a tertiary cancer centre. EQ-5D-5L utilities were derived using Australian preference weights. Interaction terms were introduced stepwise, and non-linearities were assessed using generalized additive models. Five candidate models were evaluated. Four regression frameworks were compared: ordinary least squares (OLS), Tobit (right-censored at 1), censored least absolute deviations (CLAD), and beta regression. Models were estimated with and without intercepts, with selection based primarily on root mean squared error (RMSE), mean absolute error (MAE), and information criteria (AIC, BIC). Results Beta regression with intercept provided superior predictive performance (RMSE = 0.131; MAE = 0.084; AIC = -647.3; BIC = -559.5), outperforming OLS, Tobit, and CLAD approaches. Disease-specific FACE-Q domains, including oral competence, speaking, and smiling, demonstrated strong associations with EQ-5D-5L utilities. Models with intercepts consistently outperformed no-intercept specifications across all regression methods. Conclusion This study provides the first validated mapping algorithm from FACE-Q Head and Neck Cancer Module to EQ-5D-5L, enabling economic evaluation of jaw reconstruction procedures. This algorithm facilitates cost-effectiveness analyses when direct EQ-5D-5L collection is not feasible.