Experimental control of block copolymer segregation is ideally achieved by varying temperature, but in practice, the range of segregation is very restricted. A far greater range can be accessed by diluting the block copolymer melt with a solvent, assuming the mixture obeys the so-called dilution approximation. We examine the accuracy of this dilution approximation as a function of solvent quality, size, and selectivity, using self-consistent field theory (SOFT). Naturally, the solvent quality must be good or else the solvent is prone to macrophase separation. Furthermore, the solvent size and selectivity must be sufficiently small so that the solvent distributes evenly throughout the mixture. We conclude by deriving a simple formula to specify the necessary constraints on the solvent size and selectivity.