Understanding population responses to environmental conditions is key in the current context of climate change and the extreme climatic events that are threatening biodiversity in an unprecedented way. In this work, we provide a framework for understanding butterfly population responses to weather and extreme climatic seasons by taking into account topographic heterogeneity, species' life-cycles and density-dependent processes. We used a citizen-science database of Mediterranean butterflies that contains long-term population data (28 years) on 78 butterfly species from 146 sites in the Mediterranean mesic and alpine climate regions. Climatic data were obtained from 93 meteorological stations operating during this period near the butterfly sites. We studied how seasonal precipitation and temperature affect population growth while taking into account the effects of density dependence. Our results reveal (i) the beneficial effects of winter and spring precipitation for butterfly populations, which are most evident in the Mediterranean region and in univoltine species, and mainly affect the larval stage; (ii) a general negative effect of summer rain in the previous year, which affects the adult stage; and (iii) a consistent negative effect of mild autumns and winters on population growth. In addition, density dependence played a major role in the population dynamics of most species, except for those with long-term negative population trends. Our analyses also provide compelling evidence that both extreme population levels in previous years and extreme climatic seasons in the current year provoke population crashes and explosions, especially in the Mediterranean mesic region.