Understanding the conflict potential of climate variability is critical for assessing and dealing with the societal implications of climate change. Yet, it remains poorly understood under what circumstances – and how – extreme weather events and variation in precipitation patterns affect organized violence. This dissertation suggests that the impacts of climate variability on organized violence are conditional on specific climate patterns, the sensitivity of livelihoods, and state governance. These theoretical conjectures are subjected to novel empirical tests in four individual essays. Three essays investigate the relationship between climate variability and communal and civil conflict through sub-national quantitative analysis focusing on Sub-Saharan Africa. The fourth essay sheds light on causal mechanisms leading to participation in land-related conflict based on interview material on 75 ex-participants in violence from Mt. Elgon, Kenya. Essay I suggests that the exposure of vulnerable agricultural livelihoods to sustained drought increases the risk of civil conflict violence. Essay II indicates that rainfall anomalies increase the risk of communal violence, an effect which is amplified by political marginalization. Essay III finds support for the proposition that volatility in resource supply increases the risk of communal conflict over land and water in remote regions, which tend to have limited state presence. Essay IV proposes that individuals depending on agriculture are prone to participate in land-related conflict as they face impediments to leaving a conflict zone, and additionally have high incentives to partake in fighting for land. Taken together, the dissertation furthers our understanding of the specific economic and political context under which climate variability impacts armed conflict. This knowledge is important for conflict-sensitive adaptation to climate change and conflict prevention efforts.