BACKGROUND: Prolonged intermittent renal replacement therapy (PIRRT) is a dialysis modality for critically ill patients that in theory combines the superior detoxification and haemodynamic stability of the continuous renal replacement therapy (CRRT) with the operational convenience, reduced haemorrhagic risk and low cost of conventional intermittent haemodialysis. However, the extent to which PIRRT should replace these other modalities is uncertain because comparative studies of mortality are lacking. We retrospectively examined the mortality data from three general intensive care units (ICUs) in different countries that have switched their predominant therapeutic approach from CRRT to PIRRT. We assessed whether this practice change was associated with a change in mortality rate. METHODS: Data were analysed from ICUs in New Zealand, Australia and Italy. The study population comprised all patients requiring renal replacement therapy from 1 January 1995 to 31 December 2005 (n = 1347), the period of time spanning the change from CRRT to PIRRT in each unit. Poisson regression models were used to estimate the incident rate ratio (IRR) for death, comparing the periods before and after change to PIRRT in each unit. Estimates were adjusted for patient illness severity (APACHE II score) and for the underlying time trend in mortality rate over time. RESULTS: The change from CRRT to PIRRT was not associated with any increase in mortality rate, with an adjusted IRR of 1.02 (0.61-1.71). The IRR was virtually identical in the three ICUs (P-value = 0.63 for the difference in the IRR between ICUs). CONCLUSIONS: Switching from CRRT to PIRRT was not associated with a change in mortality rate. Pending the results of a randomized trial, our study provides evidence that PIRRT might be equivalent to CRRT in the general ICU patient.