Objectives: To investigate whether eating alone (measured objectively and subjectively) and engagement in social activities are associated with loneliness and self-rated health, independently or through interaction effects among community-living adults in Sweden. Methods: Data from a national cross-sectional survey of 695 adults aged 70 to 75 years were used to assess both the subjective experience and frequency of eating alone and engagement in social activities and their associations with loneliness, and self-rated health using logistic regression and interaction analysis. Results: Eating alone and being bothered by eating alone were both associated with loneliness, but no interaction between the two was identified, meaning that the association between eating alone and loneliness was not dependent on whether eating alone was perceived as bothersome or not. Lower engagement in social activities was also associated with loneliness. However, no interaction effect was identified between eating alone and social activities. Lower engagement in social activities was associated with lower self-rated health, but no associations between eating alone or being bothered by eating alone and self-rated health were observed. Discussion: In this sample, eating alone was associated with loneliness but not lower self-rated health. The proportion of loneliness was high in groups that were bothered by eating alone, indicating specific subgroups for whom eating alone can be problematic. More research is needed to understand the issue of eating alone – and who it affects. This will allow effective policies to be developed and implemented among groups vulnerable to loneliness and poor health.