The simple relation implied by a tree model of the domestic distribution network fits data provided by the Western States Information Network for 1984-1991 quite well for various illicit drugs. Quality premia, however, are less well-explained. Price is not a function of pure quantity alone; customers pay more for 2 grams at a given purity than they do for 1 gram at double that purity. Nevertheless, some purity premia are observed for white heroin, brown heroin, and powder cocaine, although not for methamphetamines, crack, or heroin tar. Estimated coefficients reflect known phenomena such as collapses in the price of cocaine and black tar heroin, intuitively reasonable but not documented effects such as discounts for brown heroin near the Mexican border, and some unexpected results such as an apparent difference in the distribution of sinsemilla and other cannabis products. The authors suggest that similar equations reasonably describe prices of various illicit drugs in a larger region at a variety of market levels ranging from the street up to the wholesale level. This means that drug policy researchers may need to use analytical models of price as a function of transaction size in their analysis. 13 references and 6 tables