Developing policy and strategic initiatives to increase population levels of physical activity (PA) requires constant referral to the epidemiological evidence base. This paper updates the evidence that PA confers a positive benefit on health, using research studies in the peer-reviewed scientific literature published between 2000-2003. Areas covered include updates in all-cause mortality and in cardiovascular disease prevention, diabetes, stroke, mental health, falls and injuries, and in obesity prevention. Recent evidence on PA and all-cause mortality replicates previous findings, and is consistent with current Australian moderate PA recommendations. Recent papers have reinforced our understanding of the cardiovascular protective effects of moderate PA, with new evidence that walking reduces the risk of CVD and, in two studies, at least as much as vigorous activity. The evidence base for protective effects of activity for women, older adults and for special populations has strengthened. Cancer prevention studies have proliferated during this period but the best evidence remains for colon cancer, with better evidence accumulating for breast cancer prevention, and uncertain or mixed evidence for the primary prevention of other cancers. Important new controlled-trial evidence has accumulated in the area of type 2 diabetes: moderate PA combined with weight loss, and a balanced diet can confer a 50-60% reduction in risk of developing diabetes among those already at high risk. Limited new evidence has accumulated for the role of PA in promoting mental health and preventing falls.