This paper is concerned with the determination, in a cross table, of the simultaneous merging of rows and columns that maximizes the association between the row and column variables. We present a heuristic, first introduced in [2], and discuss its reliability. The heuristic reduces drastically the complexity of the exhaustive scanning of all possible groupings. Reliability is assessed by means of a series of simulation runs. The outcomes reported show that though the quasi optima may often miss the true global optima, they provide very close solutions.