The US Public Health Service Panel on Cost-Effectiveness issued a series of recommendations designed to improve the rigor and consistency of cost-effectiveness analyses. While the Panel's individual recommendations are largely sound, they nevertheless vary in importance. That is, the violation of some recommendations will yield dramatically different cost-effectiveness estimates and resource allocation decisions than the violation of other recommendations. OBJECTIVE: The Panel has advocated the use of quality-adjusted life-years (QALYs) as the best way to evaluate outcomes in a cost-effectiveness analysis. We consider the importance of this recommendation for cancer prevention, screening, and treatment by studying the empirical relationship between cost/life-year and cost/QALY. In addition, we consider whether adjusting for health-related quality of life (QOL) affects the ultimate resource allocation decision implied by the cost-effectiveness ratio. METHODS: We identified 198 articles reporting two or more outcome measures for the same intervention: cost/life-year, cost/QALY, total life-years, total QALYs, incremental life-years, or incremental QALYs. We calculated a correlation matrix for these outcomes and performed a regression analysis to examine the relationship between cost/life-year and cost/QALY. We also employed various willingness to pay (WTP) thresholds to assess whether the use of cost/life-year would yield different resource allocation decisions than cost/QALY. RESULTS: The correlation between the total life-years and total QALYs associated with the intervention is 0.97 (P<0.0001). The correlation between cost/life-year and cost/QALY is 0.78 (P<0.0001). Assuming a $50,000 WTP threshold, adjustment for QOL would affect choice in 7% of cases. With a $400,000 threshold, QOL would affect choice in 2% of cases. CONCLUSION: The outcome measures of life-years and QALYs are highly correlated with one another. While adjusting for QOL can make an important difference in some economic analyses, it generally does not affect implied resource allocation decisions for cancer prevention, screening, and treatment.