Aging is associated with difficulties in emotion identification. The underlying mechanisms of this age-related decline are not well understood, but functional brain changes may be involved. Oxytocin (OT) modulates emotion identification, but this is understudied in aging and utilizing chronic administration. To examine the effects of chronic OT on emotion identification, 107 older adults self-administered 24 IU of OT (N = 56, M = 72.8 years, SD = 7.5, 28.6% female) or a placebo (P) (N = 51, M = 70.3 years, SD = 7.4, 25.5% female) twice-daily for four weeks in a randomized, between-subject, double-blind design. Participants identified emotions via different modalities (audio, video, audio+video) before and after OT/P. Emotions were categorized by valence and arousal. Inconsistent with predictions, there were no significant treatment effects on accuracy for valence across and within modalities. However, there was a significant interaction between treatment, arousal, and timepoint across all modalities (F(1,105) = 5.11, p = 0.03, η2p = 0.05). Accuracy for low arousal emotions was greater in the OT group at post- compared to pre-intervention (p = 0.001). This effect did not hold for high arousal emotions or in the P group (ps > 0.05). Inconsistent with predictions, region-of-interest (ROI-to-ROI) analyses showed that chronic OT (>P) was not associated with altered resting-state functional connectivity of the amygdala or medial prefrontal cortex at post-intervention (p-FDR corrected > 0.05). Exploratory analyses revealed that chronic OT (>P) largely reduced anterior cingulate cortex (ACC) connectivity with other ROIs at post-intervention. Increased anterior insula (AI) connectivity with the supplementary motor area was furthermore related to improved accuracy for low arousal emotions after chronic OT (>P) across all modalities. While chronic OT may not generally enhance emotion identification accuracy in aging, OT may improve accuracy for more ambiguous emotions (i.e., low arousal) and by modulating salience network connectivity.