Managing asset integrity is crucial for the cost-effective asset management of processing plants. Consequently, new techniques for detection and classification of incipient faults are continuously being sought by industry and developed by research and development professionals, both corporate and academic. Although acoustic emission (AE) testing for faults in static equipment has been used since the 1970s, its use as a monitoring technology for various machinery conditions has been poorly adopted by industry, despite a significant volume of work having been published over the past twenty years describing success in detecting numerous rotating and reciprocating machinery faults. Anecdotal evidence from industry suggests that many 'tried and failed'. The authors believe that this is because applying AE monitoring to industrial plant is fraught with poorly documented challenges, obstacles, and limitations that must be well understood and overcome before any reported results can be replicated. Thus, to enable potential users to approach AE monitoring with more realistic expectations, the current paper discusses several such problems and suggests some available techniques for their management. After an initial introduction on the basics of AE specifically applied to machinery monitoring, issues are divided into generic problems and application-specific challenges.