Objective To identify which specific depressive symptoms predict remission to aripiprazole augmentation in late‐life treatment resistant depression. Methods This is a secondary analysis of data from a late‐life treatment resistant depression trial examining the safety and efficacy of aripiprazole augmentation. Participants aged 60 and above were randomized to aripiprazole augmentation (N = 91) versus placebo (N = 90). The main outcome was depression remission. Clinical predictors included individual Montgomery‐Asberg Depression Rating Scale (MADRS) item scores categorized as symptomatic (scores >2) or nonsymptomatic (scores ≤2). Results Three MADRS items predicted depression remission with aripiprazole augmentation: symptomatic scores on sleep disturbance and nonsymptomatic scores on apparent sadness and inability to feel. The 2‐way and 3‐way interaction terms of these MADRS items were not significant predictors of remission; therefore, the models' ability to predict remission was not improved by combining the significant MADRS items. Conclusions The identification of specific depressive symptoms, which can be clinically assessed, can be used to inform treatment decisions. Older adults with treatment resistant depression that present with sleep disturbances, lack of apparent sadness, or lack of inability to feel should be considered for aripiprazole augmentation.