The effective point of measurement (EPM) of a Kemp-Barber parallel plate ionisation chamber exposed to cobalt-60 has been determined experimentally. The variation of the EPM as a function of plate separation and the build-up cap thickness has also been studied. In general, for a constant size of build-up cap, the EPM moves downstream from the inner front wall towards the back wall of the chamber as the plate separation decreases. For parallel plate chambers, conventional theoretical analyses suggest that the EPM is the inner front wall and that it shifts towards the geometric centre of the chamber as the plate separation increases. The experimentally determined variation of the EPM, which appears to contradict these conclusions, suggests that the distribution of ionisation within a parallel plate chamber is not adequately accounted for in present theoretical descriptions. Such considerations may also affect other experimental determinations of the EPM for cylindrical chambers, as many are based on a comparison using parallel plate ionisation chambers with an assumed EPM on the inner front wall.