To check claims of a "loneliness epidemic," we examined whether current cohorts of older adults report higher levels and/or steeper age-related increases in loneliness than earlier-born peers. Specifically, we used 1,068 age-matched longitudinal reports (M-age observations = 79 years, 49% women) of loneliness provided by independent samples recruited in the German city of Berlin in 1990 and 2010, n = 257 participants in the Berlin Aging Study (BASE) and n = 383 participants in Berlin Aging Study II (BASE-II). Using multilevel models that orthogonalize between-person and within-person age effects, we examined how responses to items from the UCLA Loneliness Scale provided by observation-matched cohorts differed with age and across cohorts, and if those differences might be explained by a variety of individual factors. Results revealed that at age 79, the later-born BASE-II cohort reported substantially lower levels of loneliness than the earlier-born BASE cohort (d = -0.84), with cohort differences accounting for more than 14% of the variance in loneliness. Age trajectories, however, were parallel without evidence of cohort differences in rates of within-person age-related changes in loneliness. Differences in gender, education, cognitive functioning, and external control beliefs accounted for the lion's share of cohort-related differences in levels of loneliness. Results show that loneliness among older adults has shifted to markedly lower levels today, but the rate at which loneliness increases with age proceeds similarly as 2 decades ago. Future studies should investigate how psychosocial functioning across the life course is progressing in different sociohistorical contexts and in other age groups, such as younger and middle-aged adults.