Objectives: War-exposed refugee children are at elevated risk for mental health problems, but a notable proportion appear resilient. We aimed to investigate the proportion of Syrian refugee children who can be considered resilient, and applied a novel approach to identify factors predicting individual differences in mental health outcomes following war exposure. Methods: The sample included 1,528 war-exposed Syrian refugee children and their primary caregiver living in refugee settlements in Lebanon. Children were classed as having low symptoms (LS) if they scored below clinically validated cut-offs for post-traumatic stress disorder (PTSD), depression and externalising behaviour problems. Children scoring above any cut-off were classified as having high symptoms (HS). Each LS child was matched with one HS who reported similar war exposure, to test what differentiates children with similar exposures but different outcomes. Results: 19.3% of the children met our resilience criteria and were considered LS. At the individual level, protective traits (e.g. self-esteem; OR = 1.51, 95% CI [1.25, 1.81]) predicted LS classification, while environmental sensitivity (OR = 0.69, 95% CI [0.59, 0.82]), poorer general health (OR = 0.71, 95% CI [0.58, 0.87]) and specific coping strategies (e.g. avoidance; OR = 0.90, 95% CI [0.85, 0.96]) predicted HS classification. Social/ environmental predictors included perceived social support (OR = 1.23, 95% CI [1.02, 1.49]), loneliness and social isolation (OR = 0.85, 95% CI [0.80, 0.90]), child maltreatment (OR = 0.96, 95% CI [0.94, 0.97]), and caregiver mental and general health (e.g. caregiver depression; OR = 0.94, 95% CI [0.92, 0.97]). Conclusions: Future research should take multiple dimensions of functioning into account when defining risk for mental health problems and consider the identified predictors as potential targets for interventions.