The main objectives of this study are to discuss on the diversity of random spatial aggregation, to compares pros and cons of the different random spatial aggregation procedures, and to recommend a procedure which is believed to be superior in validity. A basic procedure for random spatial aggregation is proposed and an implementation algorithm based on the spatial proximity matrix is introduced. All six distinctive algorithms are derived on the basis of two dimensions which facilitate the diversity of random spatial aggregation. A simulation experiment is carried out to compare pros and cons of the six procedures and its results are evaluated based on the contiguousness and compactness criteria. Main findings are as follows. First, Category B seems to be more consistent in terms of contiguousness in comparison to Category A and C. Second, Category 2 appears to generate more compact final zones in comparison to Category 1. Third, Category B and C (especially B) seem to be more consistent in terms of compactness. Based on all the experiment results, Type B2 seems to be the superior algorithm for random spatial aggregation. It tends to generate more compact final zones and more consistent results in terms of both contiguousness and compactness. This study can be seen as a significant achievement in the sense that it shows that the concept of random spatial aggregation can be considered in a diversified manner, and one out of a certain group of candidates can be chosen as superior on the basis of the degree of randomness and relevancy.