Purpose: Tumor hypoxia is associated with a poor prognosis, hypoxia modification improves outcome, and hypoxic status predicts benefit from treatment. Yet, there is no universal measure of clinical hypoxia. The aim of this study was to investigate whether a 26-gene hypoxia signature predicted benefit from hypoxiamodifying treatment in both cancer types.Experimental Design: Samples were available from 157 T2-T4 laryngeal cancer and 185 T1-T4a bladder cancer patients enrolled on the accelerated radiotherapy with carbogen and nicotinamide (ARCON) and bladder carbogen nicotinamide (BCON) phase III randomized trials of radiotherapy alone or with carbogen and nicotinamide (CON) respectively. Customized TaqMan low density arrays (TLDA) were used to assess expression of the 26-gene signature using quantitative real-time PCR. The median expression of the 26 genes was used to derive a hypoxia score (HS). Patients were categorized as TLDA-HS low ( median) or TLDA-HS high (>median). The primary outcome measures were regional control (RC; ARCON) and overall survival (BCON).Results: Laryngeal tumors categorized as TLDA-HS high showed greater benefit from ARCON than TLDA-HS low tumors. Five-year RC was 81% (radiotherapy alone) versus 100% (CON) for TLDA-HS high (P ¼ 0.009). For TLDA-HS low, 5-year RC was 91% (radiotherapy alone) versus 90% (CON; P ¼ 0.90). TLDA-HS did not predict benefit from CON in bladder cancer.Conclusion: The 26-gene hypoxia signature predicts benefit from hypoxia-modifying treatment in laryngeal cancer. These findings will be evaluated in a prospective clinical trial.