Major depressive disorder (MDD) has been a leading cause of disease burden among neuropsychiatric conditions for the past 30 years. There are significant sex differences in its distribution, symptoms, and treatment outcomes, with women consistently showing higher incidence across geographic, cultural, and socioeconomic groups—suggesting underlying biological factors. Several brain regions implicated in MDD exhibit sexual dimorphism, but the nucleus accumbens (NAc) has received particular attention due to its role in reward circuitry, its sexually dimorphic features in both healthy and disease states, and its altered activity in MDD. The primary neuron type found in the NAc is the medium spiny neuron (MSN), consisting of two major subtypes defined by dopamine receptor expression (D1 or D2) and other molecular signatures. D1-MSN activity is broadly associated with reward, and D2-MSN activity with aversion, though these roles are more nuanced. In this study, we establish baseline sex differences in the transcriptomic profiles of these MSN subtypes to understand potential sources of vulnerability contributing to higher MDD rates in women. We use MSN subtype transcriptomic data from unstressed adult mice, along with NAc tissue from juvenile mice and healthy human patients. We then apply chronic witness defeat stress (CWDS), a highly translatable social defeat paradigm, to female mice to examine transcriptional alterations induced by chronic stress—the strongest predictor of MDD onset. Using differential gene expression (DEG) and weighted gene coexpression network analysis (WGCNA), we characterize transcriptional profiles of high and low social interaction phenotypes defined through the three-chamber social interaction test. Through consensus module analysis, we compare these profiles with data from male mice undergoing chronic social defeat stress (CSDS) and human MDD patients to identify a sex- and MSN subtype–specific gene coexpression module involving the PI3K–Akt–mTOR pathway. Finally, using DeepLabCut-derived social microbehavior variables from video recordings, we perform factor analysis to derive a five-factor structure of sociability and link the sex- and subtype-specific module to latent social factors to identify molecular mechanisms driving sex differences in maladaptive social behavior in MDD.