Pharmacokinetic studies of antimicrobial agents usually involve the determination of tissue distribution, since most infections develop outside the vascular compartment. Comparative analyses of the results are often complicated by their extreme variability, depending on the various types of methods used, the type of antimicrobial agent, and the characteristics of the various tissue compartments. In general, concentrations in tissue and body fluids with no barriers to penetration of drugs are fairly predictable from blood levels in relation to simple diffusion through capillary pores. This is valid for hydrosoluble drugs such as the beta-lactams as well as liposoluble drugs like the quinolones and most of the macrolides. Thus we may derive, in a predictive sense for the clinical results, both the hematic and interstitial fluid pharmacokinetic profile of the above molecules, which are useful in relation to the MIC or MBC of extracellular pathogens. For molecules such as azithromycin, in which tropism is mainly intracellular, hematic levels do not reflect those which are therapeutic in the infection site. In tissues and body fluids with barriers to the penetration of drugs, such as the central nervous system, eye and bronchial secretions, it is always useful and especially for any new antimicrobial agent to determine the concentrations which are achieved in relation to both dosage and pathological state of the tissue compartment. The significance of the chemotherapeutic formula which gives substantial predictive value of the therapeutic efficacy at the level of blood concentrations in relation to minimal inhibitory concentration of sensitive bacteria, needs to be critically re-evaluated in light of the pharmacokinetic behavior of newer antimicrobial agents.