Volunteers can be a valuable human resource for disaster risk management. In the aftermath of natural hazard events affecting vulnerable communities, there are many situations where there is an urgent need to properly match volunteers to volunteering tasks in order to fully harness their capabilities. Our study explores the relationships between volunteers' capabilities and their per-ceived satisfaction and performance in conducting volunteering tasks. We carried out a cross-sectional study recruiting a convenience sample of 170 volunteers (spontaneous and affiliated volunteers) who previously participated performing volunteering tasks during socio-natural dis -asters in Chile. We collected information on sociodemographics, experience in volunteering ac-tivities, capabilities (i.e. competencies and relatively stable psychological attributes), and per-ceived satisfaction and performance in volunteering tasks. Predictive models based on stepwise logistic regressions were fitted, with areas under the receiver operating characteristics curve (AUC) being estimated to assess the discrimination capacity of the predictive models. A heteroge-neous set of sociodemographic characteristics, experiences in volunteering activities, and capa-bilities were statistically significant predictors for each of the assessed volunteering tasks. Most of the predictive models had an adequate discriminatory capacity (AUC >0.70), with models pre-dicting performance (AUC >0.80) exhibiting the best discriminatory capacity. Knowledge of vol-unteering tasks - a competency - was the most consistent predictor of perceived satisfaction and performance in volunteering tasks. Our findings are a step forward in matching volunteers to vol-unteering tasks during emergencies from socio-natural disasters. In addition, some of the ob-served characteristics (e.g. knowledge) could be trained to improve volunteers' satisfaction and performance.