The concept of climate-smart agriculture (CSA) is gaining momentum across the globe. However, it is not specific on what should be covered under its three pillars—productivity, resilience and mitigation. Consequently, CSA encompasses many different agricultural practices/technologies, making it difficult to prioritise CSA objectives. Firstly, there is a lack of clear and workable criteria as well as methods for assessing the climate-smartness of interventions. Secondly, little information exists about the impact of the various interventions already promoted as CSA, especially in the developing world. Finally, CSA prioritisation does not take into account stakeholders’ perspectives to ensure that the interventions are applicable, suitable and of high adoption-potential. Here, we describe a new participatory protocol for assessing the climate-smartness of agricultural interventions in smallholder practices. This identifies farm-level indicators (and indices) for the food security and adaptation pillars of CSA. It also supports the participatory scoring of indicators, enabling baseline and future assessments of climate-smartness to be made. The protocol was tested among 72 farmers implementing a variety of CSA interventions in the climate-smart village of Lushoto, Tanzania. Farmers especially valued interventions that improved soil fertility and structure, reduced surface runoff, and reclaimed degraded land due to the positive impacts on yield and off-season crop agriculture. Mostly, the CSA interventions increased animal production, food production, consumption and income. The protocol is easy to adapt to different regions and farming systems and allows for the better prioritisation of interventions. But we recommend that CSA is adopted as part of a monitoring, evaluation and learning process.