Self-esteem level has been positioned as a key mediating mechanism accounting for the effects of ostracism on behaviors, invoking the notion that individuals seek to verify their self-perceptions by behaving consistent with those self-perceptions. However, evidence supporting the relation of ostracism and self-esteem level to behavioral outcomes has been mixed. We argue that such mixed effects arise because individuals may alternately seek to verify their self-perceptions via behavioral outcomes (suggesting a relation between self-esteem level and behavioral outcomes), or self-enhance via behavioral outcomes regardless of self-esteem level (suggesting no relation between self-esteem level and behavioral outcomes); the larger question is when we self-verify and when we self-enhance. To that end, we position contingent self-esteem - or the extent to which an individual bases their self-worth on outcomes in a particular domain - as a determining factor in whether we self-verify or self-enhance and present a moderated mediation model to account for varying relations between ostracism and job performance. Our predictions regarding self-verification and self-enhancement motivation are fully supported across an experimental pilot study and two field samples using multi-wave, multi-source study designs. Theoretical and practical implications for self-verification and self-enhancement motivation, as well as negative interpersonal behaviors at work, are discussed.