Urban communities face rising flood risks, highlighting the need to identify the factors that shape mitigation actions across diverse socio-spatial contexts. This study applies a configurational theory to examine how objective and subjective aspects of Flood Risk Perception (FRP) interact to jointly influence mitigation actions. Based on an online survey of 392 respondents in BouSaada, a semi-arid Algerian city, a theoretical framework was developed and validated using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results indicate that objective risk perceptions are the primary drivers of mitigation strategies, whereas subjective perceptions predominantly drive non-structural measures. Complementing these findings, fuzzy-set Qualitative Comparative Analysis (fsQCA) reveals that high mitigation levels in planned areas stem from 11 distinct configurations of objective and subjective conditions, whereas only three sufficient pathways emerge in unplanned areas. Insights from six semi-structured interviews further corroborate and clarify these patterns. The findings inform an actionable framework for optimizing context-specific interventions to strengthen urban flood resilience across diverse settings.