In Vietnam, the central Government issues land price brackets, and people's committees promulgate land price tables for five years. The published land prices are always lower than transaction prices in the market, resulting in many problems such as loss of state budget, inequality in charge of compensation, and land clearance for citizens. The study aims to uncover factors that influence land prices to build optimal land valuation models for land valuation. Using the Bayesian Model Average, the findings show that several factors strongly influence urban residential land prices in Hanoi's outskirts (Quoc Oai district), such as road type, hospitals, and industrial areas. Other factors as green space and schools, are insignificant in land purchases. The finding also indicated that the optimal model in this study showed a slight difference in valuation prices from the actual market prices. The study also determined that the mass land valuation model would be promising in Vietnam and dual-price countries. •Land prices issued by the Government lower than the transaction prices in the land use market causes unequal for citizens.•The optimal valuation model can find the specifications that best fit empirical data.•The optimal land valuation models show a slight difference in price from transaction prices.•The distance to school and green area factors minorly impact the residential land prices in suburban areas.