Acoustic emission (AE) events and event intensities (e.g., event amplitude, counts, duration, and energy counts) were recorded and analyzed during fatigue loading of uncoated and porous-coated Ti-6Al-4V. AE source location, spatial filtering, event, and event intensity distributions were used to detect, monitor, analyze, and predict failures. AE provides the ability to spatially and temporally locate multiple fatigue cracks, in real time. Fatigue of porous-coated Ti-6Al-4V is governed by a sequential, multimode fracture process of: transverse fracture in the porous coating; sphere/sphere and sphere/substrate debonding; substrate fatigue crack initiation; slow and rapid substrate fatigue crack propagation. Because of the porosity of the coating, the different stages of fracture within the coating occur in a discontinuous fashion. Therefore, the AE events generated are intermittent and the onset of each mode of fracture in the porous coating can be detected by increases in AE event rate. Changes in AE event rate also correspond to changes in crack extension rate, and may therefore be used to predict failure. AE offers two distinct advantages over conventional optical and microscopic methods of analyzing fatigue cracks--it is more sensitive and it can determine the time history of damage progression. The magnitude of the AE event intensities increased with increasing stress. Failure mechanisms are best differentiated by analyzing AE event amplitudes. Intergranular fracture and microvoid coalescence generated the highest AE event amplitudes (100 dB), whereas, plastic flow and friction generated the lowest AE event amplitudes (55-65 dB). Fractures in the porous coating were characterized by AE event amplitudes of less than 80 dB.