Traditional three-dimensional gait analyses require the skilled palpation of anatomical landmarks to identify joint parameters and produce reliable joint kinematics. Functional methods have been developed to help improve the reliability and validity of identifying joint kinematic parameters. The purpose of this study was to investigate whether a functional method could improve the between-day reliability of joint kinematics during running compared to a traditional manual marker placement method. It was hypothesised that the functional technique would result in greater within- and betweentester reliability. An eight-camera motion analysis system was used to evaluate the reliability of 3D lower extremity kinematics during running for both a functional and a manual marker placement technique. Reliability of the waveform shape, amplitude and offset of the kinematic curves was assessed using the coefficient of multiple correlation, range of motion and root mean square error respectively. The functional jointmethodology did not improve the within- and between-tester reliability in terms of kinematic curve shape, amplitude or offset compared to the manual placement technique. When experienced examiners are used to place the anatomical markers together with a lean subject sample, functional methods may not improve the day-to-day reliability of three-dimensional gait kinematics over traditional marker placement techniques.