Research findings are displayed on the associations and contribution of psychosocial factors: education, number of children, family relationship, quality of life and subjective functioning to the oldest-old persons’ survival in a ten-year follow-up period (project HECUBA, HRZZ IP-01-2018-2497). The sample consisted of 191 persons aged 80 to 97 years (on average 88 years), 73% women, residents of 13 retirement homes in Zagreb, Croatia. Questionnaire for the Oldest-Old was administered individually, as a structured interview, in 2008. The participants’ age of death was checked in 2018. Their average survival age was 92 years. Male participants were significantly more educated than female participants, and no other significant gender differences were found. Longer survival correlated with greater number of children and better quality of life. Multiple regression model showed 10% of the oldest-old survival variance explained by the set of observed predictor variables, with greater number of children as a single significant predictor of all participants’ longer survival. In the subsample of women, the same regression model explained 16.4% of the survival variance, with longer education and greater number of children significantly contributing to their longer survival. The findings suggest the need for further longitudinal research on the psychosocial factors of longevity, in order to improve the quality of life and the services for the growing number of the oldest-old persons, until the end of their life.

oldest-old persons, longevity, survival, psychosocial factors, gender differences