People living alone are a growing vulnerable population. We aim to see what kind of activities in old age (76+) may enhance survival for people living alone, and determine if these activities have different associations with survival in people not living alone, and whether these associations differ based on socioeconomic position and gender. The Swedish nationally representative SWEOLD study, 2011 and the Swedish cause of death register were used to conduct Cox regression analyses (n = 742). Incident mortality: 35.6%. In our preliminary results, participation in any activities in old age was related to reduced mortality compared to no participation. However, different types of activities seemed beneficial depending on living situation. Interacting with relatives and friends was protective among people living alone, while having hobbies and solving crosswords was protective among those not living alone. Socioeconomic position did not explain these associations. Most results were similar in women and men.