Under uncertainty, policymakers often initiate policy experiments, and policy pilots form a common and important form of policy experimenting. Effective policy pilots may be continued for policy development via scaling-up. However, the empirical evidence on the nature of such pilots and the processes and potential of their scaling-up is lacking. Using a comparative case design this chapter develops a model of policy change and applies it to the study of the design and scaling-up of policy pilots launched to deal with climate risks to the agriculture sector in India. More than two-thirds of India’s cultivable area is rainfed and the Government of India has established many pilots to address risks to agricultural production, especially in rainfed areas. Examining the features of fourteen agriculture policy pilots in India, this chapter asks the question, do the design characteristics of policy pilots influence their scaling-up and policy integration? The analysis reveals that despite their theoretical acknowledgment as an approach that enables higher risk taking and experimentation, in practice the primary contribution of pilots is found to be limited to operating as avenues for periodically updating existing policies and programmes through marginal changes in their current scope.