Quantitative research on climate variability and conflict is frequently criticized for being theoretically underdeveloped. In this article I discuss the most plausible suggested mechanisms connecting climate variability to conflict explicitly in reference to empirical testing. This approach could help solve the puzzle of how climate variability and conflict are related by highlighting how researchers can establish the key elements in the causal argument before moving on to testing it empirically. More specifically, I emphasize four key elements when evaluating each individual mechanism: first, who are the most relevant actors, second, what are the actors reacting toward (what type of climate variability), third, what conflict type is the most likely outcome, and fourth, what is the most appropriate temporal and spatial scale for each individual mechanism. Although empirical studies have moved toward more focus on theory and explicit tests of hypotheses derived from theoretical frameworks, an overview of how mechanisms are likely to manifest themselves and a discussion on how researchers can model them in analyses are missing in the research field. Adding technical fixes or new datasets to empirical testing does not automatically improve our understanding of the relationship between climate variability and conflict if the choices are not anchored in theoretical expectations.