Interpretation Bias Modification (IBM) interventions have been effective in reducing negative interpretation biases theorized to underlie depressive psychopathology. Although these programs have been highlighted as potential short-term interventions for depression, mixed evidence has been found for their effects on depressive symptoms. There is a need to examine attitudes towards training as well as individual difference factors that may impact symptom outcomes for IBM depression interventions. Seventy-two dysphoric young adults were randomly assigned to receive either an IBM targeting negative interpretation bias in personal evaluations or interpersonal situations or a healthy video control (HVC) condition. Compared to those who received HVC, participants in the IBM condition reported lower negative interpretation bias at posttreatment. No differences between conditions were found for symptom outcomes. Greater perceived treatment credibility and expectancy were associated with better treatment outcomes for both the IBM and HVC groups. Within the IBM group, a greater tendency toward assimilation with treatment scenarios was significantly associated with better treatment outcomes for both depressive and anger symptoms. This effect was unique from treatment credibility and expectancy. Pretreatment psychological reactance did not predict treatment response for either condition. Implications and future research directions are discussed.