To assess the cost of educating undergraduate medical students at the University of Texas-Houston Medical School (UT-Houston) in 1994-95 through the use of a cost-construction model developed elsewhere and adapted for UT-Houston. The cost-construction model identified the cost of the entire program as well as instructional costs (the cost of activities related to direct-contact teaching), educational costs (instructional costs plus the costs of general supervision), and milieu costs (educational costs plus the costs of research). The model predicted these costs based on student contact hours, student enrollment, full-time-equivalent (FTE) faculty and residents required, professional-activity profiles, faculty and resident salaries, and supporting resource costs. Sensitivity analysis was used to explore the effects of such factors as institutional overhead, joint products, volunteer faculty, residents, and varying assumptions about the percentage of time the faculty dedicated to direct-contact teaching. The four-year undergraduate medical program at UT-Houston had a total cost of $82,692,280. For an average class of 200 students, the annual per-student instructional, educational, and milieu component costs were $43,993, $57,370, and $90,660, respectively. The program required 201 FTE faculty members and 258 FTE residents. Sensitivity analysis revealed the financial effects of the various factors, some of which increased costs and some of which decreased costs. Although the cost-construction model is conceptually straightforward, there are several inherent complexities and aspects open to debate that must be understood in its application: for example, assumptions regarding faculty time; assumptions regarding the allocation of faculty time to various activities; the definition of those activities; the difficulty of separating costs; the difficulty of accurately defining total educational costs; and so on. However, this study provides a starting point for identifying the distinct costs associated with an undergraduate medical school program and should encourage further discussion and research.