In order to achieve sustainable transportation target, one of the transportation policy commonly adopted is to increase the use of public transport, for example, bus and train. It is, therefore, important to study the commuting behavior by bus and train of residents. Two techniques were applied to study commutingbehavior, i.e. preference functions and Moran's I spatial statistic. The commuting preference by bus was found to move towards distance maximization over time. Unlike decreasing trend for the slope preferences by bus by increasing LGA distances from the CBD, there is no clear increasing or decreasing trend for theslope preferences by train. Similar to bus, the slope preferences by train are relatively stable over time. A significant positive spatial association was identified for the slope preferences by bus for both O-D and D-O matrices, however the spatial variation in the slope preferences by train was found to be random.