It is impossible to predict when the next pandemic of influenza will occur; however, it is almost 35 years since the last pandemic, and the longest inter-pandemic interval recorded with certainty is 39 years. The next pandemic virus is likely to emerge in southeast Asia, as have two of the last three pandemic viruses. Complete global spread is likely to occur in 6 months or less, due to increased travel and urbanisation. It is likely that the usual inter-pandemic pattern of age-specific mortality will deviate temporarily towards higher mortality in younger adults. The extent to which this will happen is unclear, as the shift was extreme in 1918-1919 but less so in subsequent pandemics. Nevertheless, this may have important implications for the protection of essential workers such as health care, emergency service and military personnel. The extent to which elderly persons will be affected will depend upon previous exposure to similar influenza viruses. It is impossible to predict the likely increase in excess mortality that will occur when a new pandemic virus emerges. However, whilst mortality on the scale experienced in 1918-1919 is probably unlikely, there was a high level of mortality among those infected with the A/H5N1 virus in 1997, so it cannot be assumed that a future pandemic will be as mild as those in 1957-1958 or 1968-1969. There is likely to be more than one wave of infection and health services in most countries will be hard pressed to provide vaccines or to manage populations with clinical attack rates of approximately 25-30% and concomitant increases in demand for both primary and secondary health care services.