The emergence of drug resistance is a serious threat to the long-term virologic success and durability of HIV-1 therapy. Adherence has been shown to be a major determinant of drug resistance; however, each pharmacologic class of antiretroviral drugs has a unique adherence–resistance relationship. We develop an immunological model of the HIV-1 infected human immune system that integrates the unique mechanisms of action of reverse transcriptase and protease inhibiting drugs. A system of impulsive differential equations is used to examine the drug kinetics within CD4⁺ T cells. Stability analysis was preformed to determine the long-term dynamics of the model. Using the endpoints of an impulsive periodic orbit in the drug levels, the maximal length of a drug holiday while avoiding drug resistance is theoretically determined; the minimum number of doses that must be subsequently taken to return to pre-interruption drug levels is also established. Heterogeneity in inter-individual differences on drug-holiday length is explored using sensitivity analysis based on Latin Hypercube Sampling and Partial Rank Correlation Coefficient analysis. Extremely short drug holidays are acceptable, as long as they are followed by a period of strict adherence. Numerical simulations demonstrate that if the drug holiday exceeds these recommendations, the cost in virologic rebound is unacceptably high. These theoretical predictions are in line with clinical results and may also help form the basis of future clinical trials.