This conceptual paper will describe considerations that will be important in understanding how to begin development of good practice guidelines and draft regulatory guidance for Performance-based Outcome (PerfO) assessments of neurocognitive functioning. This is timely, as FDA plan to develop four new FDA guidance documents in support of Patient-Focused Outcome Measurement, under the PDUFA VI reauthorization. To date, the development and validation of cognitive PerfO has not usually explicitly followed FDA guidance or good practice recommendations for Clinical Outcome Assessments (COAs). However, there is a desire to include data from such outcomes in product labeling, for example Vortioxetine in Major Depressive Disorder. Understanding good practice for development and validation of cognitive PerfO is then an important goal. Cognitive PerfO exist because cognitive measurement concepts are difficult to observe and report in a quantifiable manner and mean that other COA types (ClinRO, PRO and ObsRO) may be unsuitable, or unreliable, or may in fact measure different concepts. Furthermore, precisely those circumstances under which we wish to measure cognition e.g. impairment through disease and drugs resulting in loss of insight, or to study neurobiological processes that are by their nature unobservable, can give rise to additional challenges in this regard. Whilst much of the existing 2009 FDA PRO guidance and later website information might be applied to cognitive PerfO, there are several aspects of development, validity and reliability that require unique consideration. These include content validity, which may need a greater focus on quantitative versus qualitative data, and elements of construct validity and reliability, that may employ novel data sources and different considerations for their measurement and evaluation. As with all COAs, consideration of the conceptual basis of a cognitive PerfO, the intended Concepts of Interest, and planned Context of Use provides a useful guide for their development and use.