9051 Background: Sleep disturbance is prevalent among metastatic breast cancer (MBC) patients and few studies have objectively evaluated this problem using polysomnography (PSG). Disturbed sleep negatively affects quality of life. The current study examines the relationship of mood and family structure with sleep parameters assessed over three nights of PSG (2 at-home, 1 in-lab). Methods: MBC patients (n = 103) and healthy controls (n = 27) were recruited. Patients were 57.8 years (SD = 7.7), non-Hispanic white (86.3%), and had Karnofsky ratings of at least 70%. Multiple regression analyses assessed relationships among depression, cohabitation, and sleep parameters. Results: Among patients, sleep architecture was significantly related to depression and marital status/cohabitation. Women reporting more depressive symptoms had less deep sleep: lower percentage (B = -.26, p = .01) and fewer minutes of REM (B = -.23, p = .02), lower % (B = -.25, p = .02) and fewer minutes of stage 2 sleep (B = -.28, p = .01), and more % stage 1 sleep (B = .28, p = .01). Marriage/cohabitation was positively related to sleep quality indicators: less wake time after sleep onset (B = -.20, p = .05), better sleep efficiency (B = .28, p = .01), more total sleep time (B = .30, p < .01), more minutes of REM (B = .23, p = .02), more stage 2 sleep (B = .27, p < .01), and less % stage 1 sleep (B = -.27, p = .01). Relative to healthy women, patients had more sleep stage transitions/hour (B = .21, p = .03), more awakenings (B = .18, p = .04), and more time awake after sleep onset (B = .21, p = .02). Slow wave (deep, non-REM) sleep was shorter for patients as measured by both minutes (B = -.20, p = .04) and % of stage 4 sleep (B = -.23, p = .02). Conclusions: Depressed patients or those living alone have greater reductions in deep sleep. Relative to healthy women, patients’ sleep disturbance and quality was worse. These findings demonstrate that the architecture of deep sleep is linked to depression and family structure.