In the public debate and media, it has been suggested that there currently is a "loneliness epidemic" in Western societies. To shed light on this pressing issue we investigated whether age-related changes in loneliness found in early studies also pertain to later-born cohorts, and whether mastery and self-efficacy have become increasingly important for explaining differences in loneliness. We used data from the Longitudinal Aging Study Amsterdam. Observations were of 4,880 Dutch people aged 55 and older between 1992 and 2016. We employed multilevel regression analyses incorporating birth cohort while controlling for age. Older adults were less lonely in later-born cohorts, although the effect size was small (d = .11 at Age 65 comparing Cohorts 1918-1927 and 1948-1957; and d = .11 at Age 75 comparing cohorts 1908-1917 and 1938-1947). Furthermore, we found a sharp age-related increase in loneliness; older people were lonelier than younger people (d = .83 comparing Ages 75 and 95 in Cohort 1908-1917; and d = .21 comparing Ages 55 and 78 in Cohort 1938-1947). Age effects thus clearly outweigh cohort effects in size. Increasing levels of mastery and self-efficacy across birth cohorts explain the observed cohort-related decline in loneliness. Mastery contributes similar to the effect size of partner status in the explanation of differences in loneliness. We conclude that there is no evidence of a loneliness epidemic among later-born cohorts of older adults relative to earlier-born cohorts. Also, mastery and self-efficacy are crucial to fully understanding loneliness in today's society. (PsycINFO Database Record (c) 2019 APA, all rights reserved).