Preclinical studies have demonstrated exercise improves various types of behaviours such as anxiety-like, depression-like, and cognition-like behaviours. However, these findings were largely conducted in studies utilising short-term exercise protocols, and the effects of lifetime exercise on these behaviours remain unknown. This study investigates the behavioural effects of lifetime exercise in normal healthy ageing C57BL/6 mice over the adult lifespan. 12 week-old C57BL/6 mice were randomly assigned to voluntary wheel running or non-exercise (control) groups. Exercise commenced at aged 3 months and behaviours were assessed in young adult (Y), early middle age (M), and old (O) mice (n = 11–17/group). The open field and elevated zero maze examined anxiety-like behaviours, depression-like behaviours were quantified with the forced swim test, and the Y maze and Barnes maze investigated cognition-like behaviours. The effects of lifetime exercise were not simply an extension of the effects of chronic exercise on anxiety-like, depression-like, and cognition-like behaviours. Exercise tended to reduce overt anxiety-like behaviours with ageing, and improved recognition memory and spatial learning in M mice as was expected. However, exercise also increased anxiety behaviours including greater freezing behaviour that extended spatial learning latencies in Y female mice in particular, while reduced distances travelled contributed to longer spatial memory and cognitive flexibility latencies in Y and O mice. Lifetime exercise may increase neurogenesis-associated anxiety. This could be an evolutionary conserved adaptation that nevertheless has adverse impacts on cognition-like function, with particularly pronounced effects in Y female mice with intact sex hormones. These issues require careful investigation in future rodent studies.