<h4>Background</h4>Cognitive impairment in Parkinson's disease (PD) is a key non-motor complication during the disease course.<h4>Objectives</h4>A review of detailed cognitive instruments to detect mild cognitive impairment (PD-MCI) or dementia (PDD) is needed to establish optimal tests that facilitate diagnostic accuracy.<h4>Methods</h4>We performed a systematic literature review of tests that assess memory, language including premorbid intelligence, and visuospatial domains (for tests of attention and executive functions see accompanying review) to determine suitability to assess cognition in PD. Based on in-depth scrutiny of psychometric and other relevant clinimetric properties, tests were rated as "recommended," "recommended with caveats," "suggested," or "listed" by the International Parkinson and Movement Disorder Society (IPMDS) panel of experts according to the IPMDS Clinical Outcome Assessment Scientific Evaluation Committee guidelines.<h4>Results</h4>We included 39 tests encompassing 48 outcome measures. Seven tests (different versions or subtests of the test counted once) were recommended, including four for memory, one for visuospatial domains, one for language (including three measures), and one for estimated premorbid intelligence. Furthermore, 10 tests (12 measures) were "recommended with caveats," 11 were "suggested," and 11 (15 measures) were "listed."<h4>Conclusions</h4>Recommended neuropsychological tests in memory, visuospatial functions, and language are proposed to guide the assessment of cognitive impairment and its progression in PD-MCI and PDD, and for use in clinical trials to stratify participants or as outcome measures. Novel measures being developed will need extensive validation research to be "recommended." © 2025 The Author(s). Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society. TABLE 1 Recommended neuropsychological tests including their psychometric properties Test Reliability Validity Sensitive to change Strengths Clinimetric limitations Recommendation level Memory RAVLT Good Good Sensitive to progression of memory impairment during PD course Multiple normative datasets Excellent psychometric properties Increased difficulty for PDD patients Unsuitable for screening Recommended CVLT/CVLT‐II/CVLT3 and PVLT Good Good Sensitive to progression of memory impairment during PD course Multiple normative datasets Excellent psychometric properties Includes also cued recall Increased difficulty for PDD patients Unsuitable for screening Recommended HVLT/HVLT‐R Adequate Adequate Sensitive to progression of memory impairment during PD course Multiple normative datasets Suitable for repeated testing Unsuitable for screening Recommended RBMT/RBMT II/RBMT III Excellent Excellent Sensitive to progression of memory impairment during PD course Brevity of administration Very good psychometric properties Lacking normative datasets in some languages Not available in all different languages Recommended Language BNT‐60 Good Excellent NA Multiple normative data Not suitable to detect early PD‐related language impairment Cross‐cultural differences BNT‐15 limited range for evaluating language impairment Recommended BNT‐30 Shorter but comparable to BNT‐60 BNT‐15 Very short Visuospatial function WAIS‐III and WAIS‐IV Matrix Reasoning Excellent Good Good Very often used in clinical practice Test is available in different languages Robust normative datasets in different populations Established factor structure including psychometric properties No ceiling/floor effects of the subtest Perceptual reasoning subtest is included in many PD research studies Validity of WAIS‐IV subtest based on results of WAIS‐III and WAIS‐R Validity of the perceptual reasoning index for PD is not known Different versions of the battery Recommended Estimated premorbid intelligence NART/NAART/NAART‐R High internal consistency Excellent NA Valid measure of premorbid intelligence in PD Illiterates or patients with reading disorders or significant visual impairment Sensitive to different languages and cultures Recommended Abbreviations: RAVLT, Rey's Auditory Verbal Learning Test; PDD, Parkinson's disease dementia; CVLT, California Verbal Learning Test including the revised versions CVLT‐II and CVLT3; PVLT, Philadelphia Verbal Learning Test; HVLT/HVLT‐R, Hopkins Verbal Learning Test/Revised version; PD, Parkinson's disease; RBMT, Rivermead Behavioural Memory Test paragraph recall subtest including revised versions RBMT II, RBMT III; BNT‐60, Boston Naming Test‐60; BNT‐30, Boston Naming Test‐30; BNT‐15, Boston Naming Test‐15; NA, evidence not available; WAIS‐III and WAIS‐IV, Wechsler Adult Intelligence Scale III, IV; MR, Matrix Reasoning; NART National Adult Reading Test; NAART/NAART‐R, North American Adult Reading Test‐35/the North American Memory Language Rey's Auditory Verbal Learning Test (RAVLT) Boston Naming Test (BNT‐60; BNT‐30; BNT‐15) California and Philadelphia Verbal Learning Test (CVLT/CVLT‐II/CVLT3 and PVLT) Hopkins Verbal Learning Test (HVLT/HVLT‐R) Rivermead Behavioural Memory Test (RBMT/RBMT II/RBMT III) Visuospatial Function Estimated Premorbid Intelligence Wechsler Adult Intelligence Scale (WAIS‐III/IV: Matrix Reasoning) National Adult Reading Test (NART/NAART/NAART‐R) All other reviewed tests can be found in Table S1 . However, the premorbid intelligence measures play a significant role by delineating the cognitive potential of patients and were recommended. It remains an open question whether other premorbid intelligence measures, such as the Wechsler Test of Adult Reading (WTAR) or the Test of Premorbid Functioning (TOPF) will show better discriminative potential than the NART or if these measures will be replaced by digitally assisted technology using premorbid data long before disease onset with perhaps more robust estimates and higher ecological validity. The limits of the current review must be fully acknowledged.