The diversity and resilience characteristic of the closing phases of human development are investigated by longitudinal study of change and development in about the last 30 years of the life-span. The Australian Longitudinal Study of Ageing (ALSA) is a population-based psychobiosocial and behavioural study of a cohort of 1,947 adults from Adelaide over the age of 70 years. Four annual waves of data collection started in 1992. At Waves 1 and 3, an extensive personal interview covered psychological, behavioural, social, biomedical, economic, and other contextuol variables. Participants also had the opportunity to provide psychological data on cognitive and psychosocial (self) dimensions through completion of a series of objective assessments and standard paper-and-pencil inventories. Data from these two waves is used to describe and explain changes in cognitive and psychosocial domains, particularly the complexity of functioning, potential gains and losses in memory and well-being, and the interplay amongst these aspects of functioning. Once a range of noncognitive and cognitive contextual factors are taken into account, individual differences in processing and intrapersonal resources contribute significantly and substantially to functioning in each domain, and the contribution from ageing per se is minimised. Furthermore, a “limited impact” hypothesis of late life development suggests that later declines may be a byproduct of rather large decrements in some subset of a population, with the majority showing stability. Insofar as heterogeneity increases with ageing, age per se is increasingly less able to predict functioning. This work identifies some key sources of heterogeneity that may be indicative of ageing well.