The contamination rate of phlebotomy tourniquets with meticillin-resistant Staphylococcus aureus (MRSA) was assessed, and it was determined whether this could be reduced by changes in practice or by the use of a physical barrier. Initially, the tourniquets of both preregistration house officers and phlebotomists were investigated, but as phlebotomists reported significantly more venepunctures daily, the trial continued solely with phlebotomists. Each day, the phlebotomists were supplied with a fresh sterile tourniquet, and after use, the tourniquets were swabbed and cultured. The rate of contamination with MRSA was 32 of 131 (25%) tourniquets. An audit of hand hygiene practice was undertaken and revealed that phlebotomists were performing hand decontamination inadequately between patients and wore wristwatches while working. Education comprising standard infection control methods to encourage good practice was given. After this, a polythene strip was used as a barrier by half of the phlebotomists during all venepunctures. Tourniquets were cultured and replaced daily as before. During this stage of the trial, the rates of contamination were 1 of 46 tourniquets (using a polythene strip) and 1 of 42 tourniquets (without using a polythene strip). In conclusion, phlebotomy tourniquets may be potential vectors for transferring bacteria, including MRSA. Contamination rates, and hence potential risk, can be reduced if hand decontamination is performed. This suggests that contamination of tourniquets is via phlebotomists' hands, not directly from patients' skin. Hand hygiene should be regarded as the most important method by which the spread of organisms can be reduced.