While current donor selection with clinical findings is generally effective, the imprecise nature of the assessment forces clinicians to remain on the conservative side. A reliable biological marker would assist donor selection and would improve donor organ utilization. We collected biopsies from 169 donor lungs before implantation. Expression levels of IL-6, IL-8, IL-10, TNF-alpha, IFN-gamma and IL-1beta were measured by quantitative real-time RT-PCR (qRT-PCR). Seventeen cases died within 30 days after transplantation. No donor factor was significantly associated with 30-day mortality. Univariate analysis of the 84 cases for development of the prediction model showed that IL-6, IL-8, TNF-alpha and IL-1beta were risk factors for mortality and IL-10 and IFN-gamma were protective factors. We analyzed the cytokine expression ratios of risk to protective cytokines. A stepwise logistic regression for 30-day mortality demonstrated that a model containing the ratio of IL-6/IL-10 was the most predictive (p = 0.0013). When applied to the remaining 85 cases for validation, the test of model fit was significant (p = 0.039). Using the cytokine ratio, we were able to define three risk groups with striking differences in survival (p = 0.0003). Multi-cytokine analysis of the donor lung graft with qRT-PCR shows significant promise as a strategy to biologically evaluate the donor lung prior to implantation.