Intensive care physicians perceive that there is seasonal variation in the number of admissions to critical care services. There is, however, little published evidence to support this belief. Data were therefore collected from five adjacent critical care units in the eastern region over a period of 8 years, in order to quantify any seasonal variation that may exist. Data on 16 355 critically ill patients were obtained between 1992 and 2000. Analysis showed clear winter peaks; December had a 30% higher admission rate than the quietest month, February. There was a small, but increasing, summer peak. The admission rate also exhibits an increasing linear trend, equivalent to a 6.6% annual increase in admissions per critical care bed. We conclude that there is significant seasonal variation in critical care activity, and that this is important to consider when planning services.