How does exposure to conflict events shape social trust? Research in political psychology predicts that conflict exacerbates group divisions, enhancing ingroup solidarities while simultaneously reducing outgroup trust. Experimental research has found support for these predictions, and yet measuring the impact of conflict on trust beyond the laboratory is difficult. For example, questions about the lasting salience of experimental treatments remain a challenge in the study of conflict. We develop an empirical strategy using geo‐coded individual‐level survey data from the Afrobarometer project and geo‐coded conflict‐event data. We draw spatial and temporal buffers around each survey respondent that allow us to test whether proximate exposure to conflict events correlates with lower social trust, as well as how far and long that salience lasts. We find that exposure to conflict reduces generalized and outgroup trust, as predicted. Contrary to our expectations, we find that it reduces ingroup trust. We investigate further and find that ingroup trust suffers most when respondents live in homogenous ethnic enclaves. Furthermore, we advance an argument that the effects of exposure to conflict are mitigated over distance and time. Our results indicate that the effect diminishes over both time and space.