Design decisions in geotechnical engineering typically need to be made under considerable uncertainty, both regarding present geotechnical conditions and future events occurring during the service life of the structure. To optimize the utility of societal investments, design decisions should consider the life cycle cost (LCC) and not only the construction cost. This paper investigates the applicability of Bayesian statistical decision theory to assist in this decision making. The paper illustrates the concepts with a practical example, where a geotechnical engineer considers three design alternatives for the foundation of a road embankment: pre-fabricated vertical drains with a surcharge, end-bearing and floating dry deep mixing columns. The effect of a potential extreme groundwater drawdown event on the LCC of these alternatives is analyzed and discussed. Concluding remarks are made on the relevance of such design tools in a structured risk management in geotechnical engineering projects.