The accuracy of spatial disaggregation techniques largely depends on their underlying density assumptions and the quality of the data applied. This paper presents the results of a comparative investigation of four spatial disaggregation methodologies to determine their relative accuracies. These methodologies include binary dasymetric, a regression model, a locally fitted regression model and threeclass dasymetric, each of which provides different solutions for explaining spatially heterogeneous density when population data is spatially disaggregated. In contrast to previous studies, we apply the spatial disaggregation techniques to a comparably larger and more varied geographical area which allows the spatial disaggregation techniques to be more rigorously tested. Results indicate that the three-class dasymetric technique generates higher levels of accuracy compared to the other spatial disaggregation techniques and this result is more conclusive than previous findings.