The purposes of this secondary study were to explore the case detection phenomena of elder abuse by determining the congruence between nurses' assessments of abuse and elders' self-reports of abuse; to identify factors that may account for differences between abusive situations and nonabusive situations; then to describe differences between abused elders correctly identified and abused elders incorrectly identified by nurses. Both quantitative and qualitative data were used. The theory of attribution directed this research. The conceptual framework consisted of four concepts: structural factors, relationship factors, elder factors, and caregiver factors. A descriptive-comparison design was used to address the research questions. The sample included 48 elder-caregiver dyads, of whom 24 were self-reported abused elders and 24 self-reported nonabused elders. Descriptive analysis was used, including chi-square and t-tests. Results indicated that the nurses' assessments of elder abuse and elders' self reports of abuse were congruent in only one-fifth (N = 5) of the abused cases (N = 24). The findings confirmed allegations that nurses have difficulty identifying elder abuse unless outright battering is observed. Five variables were significant between abused and nonabused elders, and 10 variables were significant between abused elders correctly identified by nurses and abused elders incorrectly identified by nurses.