The cubic illuminance concept has long been proposed to indicate light modelling in three-dimensional space. An issue relatively less discussed with regard to its measurement is the potential error due to the finite size of the cube centred at the reference point, yielding a parallax effect. In short, the measured cubic illuminance around a finite-sized object will differ from the designed values that are based on the assumption that the object is a point in space. This paper therefore aims to determine the frequency distribution of errors in estimating scalar ( Esr) and cylindrical ( Ecl) illuminances, vector to scalar illuminance ratio, and cylindrical to horizontal illuminance ratio, due to finite cube size. General uncertainty principle in measurement is employed by introducing random values of cube length and its spatial position. A linear trend is observed between cubic illuminance on the finite cube and the corresponding true values. The Esr and Ecl are approximated more accurately in the case of a point source with a small beam angle. The cube length also influences the accuracy of the results; larger cube length tends to yield less accurate estimations. To achieve maximum error of 20% in estimating Esr and Ecl for a given source–reference point distance, the cube length should not exceed 15% of such a distance.