Forest ecosystem services are critical for human well-being as well as functioning and growth of economies. However, despite the growing demand for these services, they are hardly given due consideration in public policy formulation. The values attached to these services by local communities are also generally unknown in developing countries. Using a case study of the Mau forest conservancy in Kenya, this study applied a choice experiment technique to estimate the value attached to salient forest ecosystem services by forest-adjacent communities. The choices were generated from an efficient design, and three models (conditional logit, random parameter logit model and random parameter logit model with interactions) were applied to the resultant data. The results revealed high levels of preference heterogeneity across households, including preferences for programs that guarantee improved forest cover, reduced flood risk, and high drinking water quality and quantity. There was a demonstrated welfare loss from choosing alternatives with medium rather than low wildlife population. Further, the results revealed that forest adjacent communities mind more about the welfare of downstream communities, as revealed by the high willingness to pay for flood mitigation, showing that these communities are not only concerned with private benefits, but also the welfare of society. Policy recommendations are also highlighted.