Structure-function activity relationships of drugs that bind to certain membrane-associated receptors must take into account the local membrane bilayer environment where the binding event occurs. The partitioning of drugs in an isotropic two-phase bulk solvent system such as octanol/buffer apparently is not a good model for drug interaction with the lipid bilayer of membranes. Knowledge of these membrane-based partition coefficients then necessitates reanalysis of other physical, chemical, and functional parameters. In this Perspective, we have reexamined the model used in the equilibrium dissociation constant (Kd) determination for certain lipid-soluble drugs based on recent experimental data describing the interaction of these drugs with the membrane bilayer. Because several lines of experimental evidence suggest that some lipophilic drugs bind to hydrophobic, intramembrane receptor sites via the membrane bilayer, the concentration of such drugs in the membrane bilayer compartment in equilibrium with the receptor needs to be considered for Kd calculations. In other words, instead of expressing the "free" and "bound" concentrations of the drug in terms of a total aqueous volume (moles of drug per liter of solution), these quantities should be expressed as a function of the membrane lipid volume (moles of drug per liter of membrane lipid). The results of this analysis indicate that Kd values calculated on the basis of an aqueous concentration of the drug are significantly different from those using the "membrane concentration" of the drug, as measured experimentally. This difference in the Kd values is related to the membrane partition coefficient of the drug. In addition to affinity constants, drug interaction with the membrane should be considered for other pharmacological parameters such as pICs and association rate constants. These parameters are important considerations for designing new therapeutic agents that have a dominant interaction with a cell membrane and a specific component of a cell membrane.