This paper estimates the impact of natural events on the economic vulnerability of Peruvian households, with a focus on the role of frequency and severity. We combine detailed, high-frequency administrative data from emergency records at the district level provided by the National Institute of Civil Defense (INDECI) and nationally representative household survey data to construct householdlevel exposure measures. We define frequency as the number of distinct weeks with recorded events in a household's district of residence, while severity is based on the number of individuals affected and displaced by each event. To account for the potential effect of treatment lags, we employ a recent differences-in-differences estimator, an approach not yet widely applied in the natural disaster literature and particularly relevant in contexts where such events are frequent and varied. While natural events overall show limited effects on economic outcomes, we find that high-frequency and high-severity exposure is associated with slower income and consumption growth, with effects that persist and even intensify over time. These findings suggest that households display a degree of resilience that weakens when events are too frequent or severe. We also document heterogeneity by type of natural event: low temperature episodes reduce income growth, whereas precipitation-related events might have positive economic effects.