Researchers have created numerous measures of social and emotional competence (SEC). These measures tend to correlate with each other and with measures of stressful events (SEs), suggesting potential redundancy. We evaluated which, if any, SEC variables predicted unique variance in social and mental health after controlling for other SEC variables in the study and the impact of SEs. 331 university students participated in an anonymous, cross-sectional study. We measured SEs, and a wide variety of SECs, including: social problem solving skill (effective problem orientation, automatic processing, and problem solving), alexithymia (difficulty identifying and describing emotions; minimising emotions), effective emotional control (low rumination, high impulse control, high aggression control, low defensive inhibiting of emotions), and level of emotional awareness. We also assessed a variety of aspects of social and mental health (depression, anxiety, hopelessness, suicidal ideation, life satisfaction, social support). Analyses revealed that all SEC measures except minimising emotions had significant incremental value over the other measures and over SEs in predicting social and mental health. The optimal set of predictors differed depending upon the type of health predicted. These findings have important implications for the design of social and emotional intervention programs.