This study aims to develop a decision making process which is more reliable and effective used for assessing and selecting alternative methods of subgrade improvement of a road on soft soil deposition. The process starts from identification of site, geomechanic characterization, finite element modeling and simulation, and selecting alternative methods by using a decision tree with a single stage decision process. Decision was established based on the quantification of probability of successful and probability of failure of each alternative method. The method with high probability of successful and low cost would become more preferable rather than that with high probability of successful with high cost, and that with high probability of failure with low cost. To show the effectiveness of the process, it implements in the case of road embankment on soft soil in Tikke - Baras Road, a section in a national road of Makassar - Palu via North Mamuju, located in West Sulawesi. Stratigraphy based on log bore data shows that the embankment overlays a clayey sand with a 9 m thick and 1 m of peat soil. Below those layers, it was found loose sand with a 12 m thick. Finite element model was conducted in which four alternative methods simulated including embankment with improved drainage, wooden pile (cerucuk) and geogrid, micropile and geogrid, and pile slab. The performance ofeach alternativemethodrelated to cost, settlement, probability of failure and additional cost was evaluated. It was found that alternative of embankment without any reinforcement would lead to high additional cost due to high probability of failure, while alternative pile slab would have consequence of high cost of construction with low probability of failure. While future research is still recommended in investigating spatial variability effect on the decision making, this study would contribute in enhancing infrastructure development on soft soil deposits.