Far too often, the basis for selection of a mud system is the "latest, greatest" technology or personal preference rather than sound cost-effective analysis. The use of risk-vs.-cost decision analysis improves mud selection and makes it a proper business decision. Several mud systems usually are available to drill a well and, with good decision analysis, the cost-effectiveness of each alternative becomes apparent. The Nigerian Natl. Petroleum Corp. and Mobil Producing Nigeria (MPN) recently began development of the Oso field, a condensate development funded by the World Bank. The budget authorized $187 million for development drilling of 19 wells and completion of 21 wells. Early drilling was very troublesome owing to wellbore instability, lost circulation, and stuck pipe. In addition, the mud cost was excessive because of high dilution rates and expensive mud products (especially barite, which was approximate to 50% of the final mud cost). This paper describes how the drilling team used structured decision analysis to evaluate and select the best mud system for the project. First, Monte Carlo simulations forecast the range of possible results with each alternative. The simulations provided most-likely values for the variables in the decision tree, including reasonable ranges for sensitivity analyses. This paper presents and discusses the simulations, the decision tree, and the sensitivity analyses. The analysis compares several factors in six mud systems, including cost-per-barrel; dilution; trip time; rate of penetration (ROP); stuck pipe, lost-circulation, wellbore-stability, torque-and-drag, filter-cake, and carbonate problems; and environmental concerns. The analysis used variables based on actual data whenever possible. The results indicated that, in spite of necessary rig modifications and increased risk of lost circulation, a mineral-oil-based mud was the best choice. Before introduction of the mineral-oil-based mud, a steady improvement had occurred in drilling performance, which improved dramatically after this mud was introduced. During the project, 21 wells were completed and 22 development wells were drilled (three more than originally planned). Total cost was $176 million, $28 million under the original budget and $11 million under the final budget. This paper presents field results that show the improvement in well, mud, and coring costs; average ROP; and trouble time. Through sound application of decision analysis, selection of the mud system becomes a sound business decision.