*sigh* yes, i'm NOT saying that it doesn't cost money to have grad students. The question is over the COST of an MD/PhD program to a school and to the government. A given school needs to have X number of people working in its labs to get research done. Grad students are the cheapest labor (trust me, as a tech I cost way more than any grad student including the cost of tuition). This money is usually covered by grant funding (direct or indirect). Now, you have X number of people that need to work at the graduate student level. They can be either in an MD/PhD program or in a PhD program. During the graduate school years of the MD/PhD program, the University is incurring the same cost for its MD/PhDs as it does for it's PhDs and getting the same benefit in terms of being one of X workers. Therefore, when examining the marginal cost of having an MD/PhD program to a university, you should exclude the fixed cost of labor during the PhD years because it will be incurred regardless of whether the students doing the work are MD/PhD or PhD only. Finally, in order for a university to survive, the average lab HAS to be bringing in more money in grant money than it is spending on personnel (tuition and salaries) to survive (through indirects that don't kick back to the labs). That means that your average grad student must be part of a team that is bringing in more money than it costs (aka a profit). Therefore, the cost to have grad students at a university that is successful is, in fact, negative. So, in conclusion, when looking at the cost of an MD/PhD program, you need only to look at the cost of the medical school years. As to cost to the gov't, it's likely only a small fraction of the cost to universities.