Our aim was to compare displayed pressure with actual intra-articular pressure using three fluid delivery systems. The pumps used were the Arthrex, FMS and Dyonics. We investigated thirty patients undergoing shoulder arthroscopy. Patients with a previous disruption to the joint capsule were excluded. A standard set-up was used with the patient in a lateral position and arm in traction. The arthroscope was introduced from the posterior portal. The rotator interval was identified and the needle introduced from this point. The needle was attached to a arterial pressure transducer. The pressure transducer was coupled to the anaesthetic machine. Pressures of 30, 50 and 80 mmHg were dialled on the fluid management systems. Once the pressures stabilised on the pumps the intra-articular pressure was measured independently using the pressure transducer. Median pressures for the FMS and Arthrex pumps were approximately twice set pressures. Median pressures for the Dyonics pump were closer to set pressures but the range of values was wider. The three pumps were inaccurate and behaved differently to each other. They are not interchangeable. Each pump performed inconsistently.