Drug-loaded polymers and polymeric microparticles provide an attractive form for controlled drug-delivery systems. Design of new systems requires knowledge of polymer-drug interactions. The effect of polymer architecture and chemistry upon active-ingredient loading is investigated by Monte Carlo simulation. The ensemble-growth method is used to sample conformations of a model polymer comprising polar and nonpolar segments. The polymer is a block copolymer, linear or branched. In our calculations, the polar portion of the polymer contains 21 segments. The polymers are dissolved in either of two types of solvent models, In the first, nonpolar solvent, the polar segments tend to collapse, but the bulky nonpolar groups, easily soluble in the medium, create some cavities in the polymer. These cavities are suitable hosts for the slightly polar active ingredient. In the second solvent, polar, the nonpolar segments contribute to attract the active ingredient within the polymer segments, therefore lowering the burst-release rate. The relative uptake of the active ingredient, proportional to the probability of finding an active ingredient within the radius of gyration of the polymer, is computed as a function of the number of nonpolar segments in the polymer. Simulation results are reported for active ingredients of two different sizes. For given size of the polar portion, short nonpolar tails increase the active-ingredient relative uptake in both solvents considered. Linear block copolymers look promising for obtaining higher entrapment efficiency for the active ingredient and for controlled release.