In line with the availability of an increasing array of systemic antifungal agents, there is a need for accurate, reproducible and predictive susceptibility testing of fungal isolates in order to help inform clinical choice. Much early attention on antifungal susceptibility testing focused on defining test parameters that produced reproducible and reliable intra- and inter-laboratory results and there are now standardized methods for the testing of yeast and mould isolates. The application of this standardized approach produces susceptibility results that are comparable between laboratories and allow epidemiological analyses at the national and even international level. In addition, monitoring and prediction of emerging susceptibility trends are now possible. As methodology improved, attention shifted to clinical outcome data in order to establish breakpoints for antifungal agents. However, there are a large number of confounding factors that must be considered when trying to assess the in vivo activity of antifungal agents in invasive disease. An approach based on defining normal susceptibility ranges, together with pharmacokinetic and pharmacodynamic analyses and identification of resistance mechanisms, has helped to establish breakpoint data for at least some of the systemic agents active against yeast isolates. Moreover, for some drugs, there is a supporting in vitro-in vivo correlation available from studies of clinical efficacy. Application of susceptibility testing has helped to define the spectrum of activity of all the currently available antifungal agents. Both intrinsic and emergent antifungal drug resistance are encountered and the predictability of innate resistance has meant that species identification is often sufficient to alert the clinician to the likelihood of in vitro and often-associated in vivo resistance. Currently, emergence of resistance in a previously susceptible strain during a course of treatment remains rare.