To determine if estrogen would protect treated rats from deficits in performance on a working memory task across time, 18 female 6-month-old Sprague-Dawley rats were trained to a criterion on a water-escape spatial delayed matching-to-sample problem. Following training, rats were ovariectomized, and nine were maintained on estrogen (polyestradiol-phosphate, 0.5 mg every 3 weeks) and nine on its vehicle for 200 days. After recovery from surgery, the rats were tested for performance every 6 weeks under three conditions: 5 min retention interval (RI); 30 min RI; and 30 min RI with an emotional experience during the RI. Analysis of correct choices revealed that estrogen-treated rats made more correct choices (p < .05) than controls on the 5 min undisturbed interval; estrogen tended to impair performance on the emotionally distracting interval. Estrogen apparently protected working memory on the undisturbed trials and might be pertinent to the maintenance of memory in female mammals.