Appropriate fluid replacement is an essential component of trauma patient resuscitation. Once haemorrhage is controlled, the restoration of normovolaemia is a priority. In the presence of uncontrolled haemorrhage, aggressive fluid resuscitation may be harmful. The crystalloid-colloid debate continues, but existing clinical practice is more likely to reflect local biases and dogma rather than evidence-based medicine. Colloids vary substantially in their pharmacology and pharmacokinetics and the experimental findings based on one colloid cannot be extrapolated reliably to another. In the initial stages of trauma patient resuscitation, the precise fluid used is probably not important, as long as an appropriate volume is given. Later, when the microcirculation is relatively leaky, there may be some advantages to colloids such as hydroxyethyl starch. Hypertonic saline solutions may have some benefit in patients with head injuries. A number of haemoglobin solutions are under development but one of the most promising of these has been withdrawn recently. It is highly likely that at least one of these solutions will eventually become routine therapy for trauma patient resuscitation. In the mean time, contrary to traditional teaching, recent data suggest that a restrictive strategy of red cell transfusion may improve outcome in some critically ill patients.