Purpose: The purpose of this study is to suggest a new approach to identifying patterns of comorbidity and multimorbidity. Design and Methods: A random sample of 1,039 rural community-resident American Indian elders aged 60 years and older was surveyed. Comorbidity was investigated with four standard approaches, and with cluster analysis. Results: Most respondents (57%) reported 3 or more of 11 chronic conditions. Cluster analysis revealed a four-cluster comorbidity structure: cardiopulmonary, sensory-motor, depression, and arthritis. When the impact of comorbidity on four health-related quality of life outcomes was tested, the use of the clusters offered more explanatory power than the other approaches. Implications: Our study improves understanding of comorbidity within an understudied and underserved population by characterizing comorbidity in conventional and novel ways. The cluster approach has four advantages over previous approaches. In particular, cluster analysis identifies specific health problems that have to be addressed to alter American Indian elders' health-related quality of life.