Abstract Utilising data from the ASPirin in Reducing Events in the Elderly trial participants [aged >70-years (Australians), >65-years (US)], we estimated percent change in body size (BMI: body mass index, WC: waist circumference) from baseline and the second annual visits. Both BMI and WC changes were categorised as 1) change within 5% (stable), 2) decrease by 5–10%, 3) decrease by >10%, 4) increase by 5–10%, and 5) increase by >10%. Mortality events were classified according to the underlying cause by adjudicators. Hazard ratios were calculated to compare mortality between body size change categories. Amongst 16,773 participants, 1,269 mortality were observed over an average 4.4±1.7 years. In men compared to stable BMI, a 5–10% decrease in BMI had a 35% higher (HR 1.35, 95% CI 1.08–1.67), and a >10% decrease in BMI had a 3.84-fold higher (HR 2.99, 95% CI 2.44–3.68) risk in all-cause mortality. A decrease in BMI was associated with a higher cause-specific (cancer, cardiovascular disease [CVD], and non-cancer non-CVD) mortality. A decrease in BMI was associated with higher mortality risk in women, however, the magnitude of association was weaker than men. A decrease in WC was predictive of mortality but with a much weaker relationship than a decrease in BMI. Weight gain was not predictive of mortality except for non-cancer non-CVD mortality in women. Physicians should be aware of the ominous consequence of weight loss, especially amongst older men. The risk extends beyond an increased risk of cancer, extending to CVD and a range of other life-limiting conditions.