High morale could be considered to be an essential part of aging well and increased knowledge of how to prevent a decreasein high morale in very old age could have important implications for policy, and social and health care development. Theobjective was to identify social and health-related risk factors for a decrease in morale over 5 years in very old peopleamong those with high morale at baseline. The study is based on data derived from the Umeå85+/GERDA study conductedin Northern Sweden and Western Finland. The final sample consisted of 174 individuals who were 85 years and older atbaseline and who had completed the follow-up 5 years later. Morale was measured with The Philadelphia Geriatric CenterMorale Scale (PGCMS). A set of social and health-related variables were used to test which factors were associated with adecrease in morale over 5 years. Linear regression was used for the multivariable analyses. The sample had a mean changeof − 1.3 (SD = 2.5) in PGCMS scores from T1 to T2. The results from the regression analyses showed that development ofdepressive disorders, increased feelings of loneliness and the death of a child during the follow-up period were associatedwith a decrease in morale. The results from our study indicate that preventing the development of depressive disorders andincreasing loneliness are key factors in preventing a decrease in high morale. Additionally, very old people who have recentlylost an adult child should receive adequate psychosocial support