Women were interviewed four times over the first five months after they left a battered women's shelter in this longitudinal experimental design. Women in the experimental condition received intensive, one-on-one services with trained paraprofessional advocates for a period of 10 weeks. Advocates assisted women in gaining access to needed community resources. Women in the control condition received no addition services. Findings corroborated earlier research which suggested that battered women often lack many resources on leaving their abusive partners. Working with an advocate resulted in women reporting being more effective in obtaining desired resources. However, whether this effectiveness influenced a woman's ability to remain free of her assailant was unclear, as the vast majority of women in both conditions were still not with their assailants at the followup assessment period (five months after exiting the shelter). 3 tables and 28 references (Author abstract modified)