This paper argues that opportunities for reducing poverty, improving social justice, and influencing policy and institutional changes are being missed as a result of not learning enough from situations where positive changes in development indicators have already taken place. After a review of research on monitoring and evaluation (M&E), positive deviance and studies of ‘success' stories, three case studies of positive change are presented. These are the spread of bamboo tubewell irrigation in Bihar, changes in agricultural research and extension policy in Nepal, and the spread of groups and group-based organizations/federations in Nepal. General lessons include: (1) effective institutional innovation is always new and social arena specific, (2) many opportunistic social entrepreneurs are always present in arenas of positive social change, and (3) there is always purposive selection of what to observe, what to measure and what to publicize. Practical implications include: (1) strengthening social science research on understanding change in agricultural and natural resources systems, (2) selecting people for research and development intervention situations based not only on technical competence but also a track record of interest in social justice development principles, and (3) strengthening a broader concepts reflection and learning within current research and development interventions chan is currently practiced. While learning from the positive is a simple idea, it is always challenging to implement as it inevitably questions the histories, past explanations and perceptions of some scientists and development planners, especially those who promote a mainstream, formulaic approach to the design and promotion of best policies and best practices, and a simplistic, non-political/cultural approach to the transfer and scaling out and up of technology and institutional models.