In view of the Indian Government's growing commitment to forestry, a number of afforestation projects have been implemented. But most projects in developing countries do not conclude as per plan, uncertainty being a major factor. This study undertakes physical, financial, economic and social appraisal of the afforestation programme through case studies of the farm forestry (FF) and the rehabilitation of degraded forest (RDF) components and discusses the conceptual and methodological issues in appraisal of these projects under risk, threat and uncertainty. We have used three different approaches to risk appraisals in the present study. They are: the expected value of NPV through the illicit felling models; the cumulative distribution function comparisons through stochastic efficiency rules; the utility function of the project managers. Physical (logistic and Weibull models), financial and management decision (deterministic and probabilistic models) models developed in the study help in threat appraisal through quantification of physical loss, financial appraisal of its consequences and formulation of a management strategy under the threat of illicit felling. The risk analysis of the FF and the RDF component using Monte Carlo simulation is used to generate probability of return profiles and the results are compared through stochastic efficiency rules. The utility functions of the project managers are used to describe their risk attitude. The study shows that most managers are risk averse and the analysis of their utility functions supports the decreasing absolute risk aversion hypothesis. It emphasizes the need for a risk policy in the Forest Department. The economic appraisal examines the interaction of the FF and the RDF components with the economy rather than the treasury. Illicit felling is accounted for as a benefit to the economy. A 'Shadow pricing approach' is adopted for economic and social appraisals. For the social appraisal, inputs and outputs are estimated in terms of net discounted utility-weighted consumption flows. All the parameters of social and economic appraisal such as the consumption value of unit reinvestment, utility weight for incremental consumption at different consumption levels, social discount rate and economic discount rate are estimated. To study farmers' adoption behaviour, principal component analysis is used to explore significant factors and a logit model is developed after that to estimate probability of adoption. The study indicates that adoption of FF can be explained in an overall framework of evolutionary theory proposed in this study. The evolutionary theory posits that farmer tree growing can be considered as a land use strategy in response to both changing macro and micro factors, many of which relate to characteristics of the farmers, their resource endowments etc. It is concluded that success of projects can be assessed by taking account of the factors influencing the variability in the project outcome and understanding the whole process of people's interaction and participation in the project.