OBJECTIVE: To estimate the range of poly(adenosine diphosphate [ADP]-ribose) polymerase expression in serous ovarian cancers and to determine whether expression is associated with response to therapy and outcome. METHODS: Immunostaining for poly(ADP-ribose) polymerase was performed in 186 paraffin-embedded, serous ovarian cancers. Nuclear poly(ADP-ribose) polymerase expression was quantified using a scoring system that assesses both staining intensity and percentage of cells staining. Kaplan-Meier analysis was performed to evaluate the relationship between poly(ADP-ribose) polymerase expression and overall survival. RESULTS: High poly(ADP-ribose) polymerase expression was present in 54% of serous cancers but was not associated with stage or grade. There was no difference in the rate of complete clinical response to primary chemotherapy between cases with low poly(ADP-ribose) polymerase expression (70%) compared with those with high poly(ADP-ribose) polymerase expression (71%). However, high poly(ADP-ribose) polymerase expression was associated with significantly worse median overall survival (36 compared with 43 months, P=.04, hazard ratio 0.71). CONCLUSION: Expression of poly(ADP-ribose) polymerase in ovarian cancers is heterogeneous, and high expression in serous ovarian cancers is associated with worse overall survival. These data suggest that evaluation of poly(ADP-ribose) polymerase expression in the primary cancer could potentially allow selective use of poly(ADP-ribose) polymerase inhibitors in patients most likely to respond. LEVEL OF EVIDENCE: III.