The Millipore filter unit has been advocated as a means of reducing the chance of bacteria entering the circulation during intravenous infusion. In a prospective study no significant reduction was obtained in the incidence of thrombophlebitis or in the bacterial contamination of cannulae. The unit was inconvenient to use and in-vitro and in-vivo studies showed reduced flow rates and frequent episodes of filter blockage. Its use was further restricted by the fact that blood and fat emulsions would not pass through it.