Activities that involve a change in direction apply a high rotational load to the knee joint. Biomechanical analysis of such activities may be useful for determining mechanisms that underlie knee injury and the success of ligament reconstruction surgery. However, the reliability of the measurement of tibial rotation remains unclear. The purpose of this study was to determine the reliability of tibial rotation measurements during a pivoting task, both between testing sessions conducted on the same day and between those made one week apart. Three-dimensional motion analysis was used to measure peak internal tibial rotation and rotational excursion during a stair descent and pivoting task in eleven healthy subjects (six female, five males). Intraclass correlation coefficients (ICC (3, 1)) and typical error analyses were used to examine within and between day reliability. Tibial rotational excursion had excellent reliability for within day (ICC=0.82) and between day sessions (ICC=0.76) whereas peak internal rotation had good reliability (within ICC=0.74; between ICC=0.68). Typical error was less than 2.4 degrees for within day measures and 2.9 degrees for between day measures. It was concluded that tibial rotation can be measured reliably during pivoting. Typical error values were less than the usual group differences in rotational excursion reported in the literature. The ability to reliably quantify tibial rotation during dynamic activities is important in determining the causes of persisting instability following anterior cruciate ligament reconstruction.