Community forestry projects in Ethiopia have been implemented using the top–down approach, which may have contributed to the failure of most of these projects. The so-called community plantations practically belonged to the government and the labour contribution of the local communities in the establishment of the plantations was mainly in exchange for wages. In this paper, we use the contingent valuation method to examine the determinants of the value of community forestry in rural Ethiopia and its feasibility, when the plantations are established, managed, and used by the communities themselves. The value elicitation format used is discrete question with open-ended follow-up which is closer to the market scenario our respondents are familiar with compared, for example, with the single discrete choice format. Unlike most other studies, we use a tobit model with sample selection in the empirical analysis of the bid function to correct for the effect of excluding invalid responses (protest zeros, outliers and missing bids) from the analysis. The analysis of the bid function shows that household size, household income, distance of homestead to proposed place of plantation, number of trees owned and sex of household head are significant variables that explain willingness to pay. We also find that there are significant differences in willingness-to-pay across sites. It is hoped that this study contributes to the limited empirical literature on community forestry in developing countries by indicating some of the conditions under which community plantations will be acceptable and feasible.