Objective-To examine the relationship between attachment anxiety, marital status, bed-partner status, and sleep in recurrently depressed women.Methods-The current study measured polysomnography (PSG) and subjective sleep quality in 107 women with recurrent major depression. Women were categorized as high or low in attachment anxiety based on Bartholomew and Horowitz's Relationship Questionnaire (1991).Results-There were no significant main or interaction effects of any of the relationship measures on subjective sleep quality. In contrast, PSG results indicated that women with bed partners displayed better sleep efficiency (p < .005). Marital status was also associated with sleep efficiency (p < .05), and married women displayed significantly shorter sleep latencies as compared with never married women (p < .05). Anxiously attached women displayed a reduced percentage of stage 3−4 sleep (p < .05). Moreover, a significant interaction between attachment anxiety and marital status (p < .05) suggested that anxiously attached women who were previously married (i.e., divorced, separated, or widowed) displayed a particularly low percentage of stage 3−4 sleep.Conclusions-Depressed women who exhibit an anxious attachment style and have experienced a marital rupture show reduced stage 3−4 sleep, which may signal a concomitant reduction in restorative cognitive and metabolic processes. Relationship context influences sleep continuity. These results provide a more nuanced approach to considering qualitative and structural aspects of relationships that may influence sleep.