In the current study the interobserver and intraobserver reliability of a recently developed method to obtain the position and orientation vectors of the flexion-extension axis of the elbow in vivo is determined. The method uses the Flock of Birds(TM) six degrees-of-freedom electromagnetic tracking device. Ten subjects performed three trials comprising five flexion and extension cycles. The movements of the forearm with respect to the upper arm were recorded. Observer A measured two trials and observer B measured one trial. Optimal instantaneous helical axes were calculated in a humeral coordinate system for each trial. Intraclass correlation coefficients and 99% confidence intervals were computed to compare the three measurements. Zero was in the range of all the narrow confidence intervals, which is strong indication for resemblence. Interobserver intraclass correlation coefficients values for orientation vectors were good to excellent and intraobserver values were fair to good. The intraclass correlation coefficients values for position vectors were lower, probably due to the lack of variance between subjects. It is concluded that the method is reliable and can be used in certain clinical settings. Copyright (C) 2000 Elsevier Science Ltd.