Smallholder livelihoods in the Peruvian Altiplano are frequently threatened by weather extremes, including droughts, frosts and heavy rainfall. Given the persistence of significant undernourishment despite regional development efforts, we propose a cluster approach to evaluate smallholders’ vulnerability to weather extremes with regard to food security. We applied this approach to 268 smallholder households using information from two existing regional assessments and from our own household survey. The cluster analysis revealed four vulnerability patterns that depict typical combinations of household attributes, including their harvest failure risk, agricultural resources, education level and non-agricultural income. We validated the identified vulnerability patterns by demonstrating the correlation between them and an independently reported damage: the purchase of food and fodder resulting from exposure to weather extremes. The vulnerability patterns were then ranked according to the different amounts of purchase. A second validation aspect accounted for independently reported mechanisms explaining smallholders’ sensitivity and adaptive capacity. Based on the similarities among the households, our study contributes to the understanding of vulnerability beyond individual cases. In particular, the validation strengthens the credibility and suitability of our findings for decision-making pertaining to the reduction of vulnerability.