Many library return on investment (ROI) studies have been conducted in the United States. Most rely on "cost savings" approaches to determine the marginal benefits of library services. These methods fail to logically have a meaningful relationship to theoretical benefits estimation. Adaptations of the contingent valuation method (CVM) to this application are discussed in reference to the well-known National Oceanic Atmospheric Administration panel guidelines and recent literature. A CVM technique is used to estimate median annual household benefits for Minnesota public library services. Results indicate that CVM likely leads to more conservative estimates than "cost savings" approaches, which is probably due to the realities of available substitutes and low patron marginal benefits from additional transactions. Evidence is provided to limit the usual concerns of the utilized methodology. Focusing on programs for low income and education households may create higher returns more directly attributable to these services.