OBJECTIVES: Loneliness arises from a discrepancy between the relationships one desires and those one experiences. This study examines the relative importance of relationship quantity and quality for loneliness in older adults. METHOD: The study was a cross-sectional, representative survey of 474 adults aged 77 and above living in Dalarna, Sweden. Loneliness was measured using the three-item UCLA Loneliness Scale. Four ordinal logistic regression models of loneliness were developed: one for quantity of relationships and three for quality of relationships (closeness
conflict
closeness and conflict combined). Perceived social support was included in the quality models. RESULTS: The quality models explained the most variance in loneliness and had better model fit. Having a partner was associated with less loneliness, even when there was conflict. Having a close relationship with children was associated with less loneliness, but a lack of closeness or the presence of conflict was associated with more loneliness. Perceived social support was associated with less loneliness. CONCLUSION: Quantity and quality of relationships are both associated with loneliness in older adults, but relationship quality appears to have greater importance. Specific types of relationships are also relevant, highlighting the need to consider relationship dynamics when designing interventions to reduce loneliness.