Despite notable progress, the management of invasive aspergillosis (IA) remains challenging and treatment failures are common. The final patient outcome is subject to multiple factors including the host (the severity of the underlying conditions), the fungus (the virulence and susceptibility pattern of the Aspergillus species involved), and the therapy (the timing related to severity of infection and choice of therapy-dose, efficacy, cidal versus static, toxicity and interaction). Consequently, assessment of failure is complex yet crucial in order to ensure appropriate management. Refractoriness in absence of drug resistance may reflect severity of the underlying disease/infection at the time of initiation of therapy prolonging time to response. It may also reflect a suboptimal antifungal drug exposure due to poor compliance, inappropriate dosing or increased drug metabolism, or it may reflect 'pseudo' failure due to worsening of imaging due to recovery of neutrophils. Refractoriness may also be related to inherent drug resistance in various Aspergillus species or acquired resistance in a normally susceptible species. The latter scenario is mostly encountered in A. fumigatus, where azole resistance is increasing and includes azole-naive patients due to resistance related to azole fungicide use in agriculture and horticulture. Although diagnostics and resistance detection have been greatly improved, the time to resistance reporting is often still suboptimal, which calls for close assessment and potentially management changes even before the susceptibility is known. In this article we address the various definitions and approaches to assessment and management of clinical refractoriness/failure in the setting of proven and probable IA.