Introduction Direct relations between social network and mortality patterns has been well documented, but less is known the relationship between size, density, boundedness of social network, social support, and psycho-social dimensions of quality of life in older age and its direct and indirect effects on all-cause mortality.

Aim The aim of the study was to assess the characteristics of social network in older stage of life in relation to health-related quality of life and its effect on mortality.

Methods The base-line study was performed in simple random sample of 552 Krakow citizen aged 65–85 yrs old, as a part of Central European Network on Healthy Ageing. Face to face structured interviews were based on Polish validated version of Dutch questionnaire. SF20 test, HADS, GARS, SSL12-I, de Jong-Gierveld Loneliness Scale and Cantril's ladder were used. Data on mortality was ascertained by monitoring city vital records and all-caused mortality was analysed. During 8 years 31.1% of study population died. Cluster analysis was performed and Cox proportional hazard model was used.

Results Cluster analysis distinguished different clusters for age group -75 yrs and 76 yrs and over. Cox proportional models revealed significantly highest risk of death (HR=2.97, 95% CI (1.21 to 7.28)) in individuals aged 76 and over who were characterised by seldom contacts with children, neighbours and others, high loneliness, low social support and low general quality of life.

Conclusions Relationship between the properties of social network and health-related quality of life remains significant predictor of mortality in advanced old age.