Epidemiological and logistical reasons are slowing the clinical validation of the molecular imaging biomarkers in the initial stages of neurocognitive disorders. We provide an updated systematic review of the recent advances (2017-2022), highlighting methodological shortcomings. Studies reporting the diagnostic accuracy values of the molecular imaging techniques (i.e., amyloid-, tau-, 18FFDG-PETs, DaT-SPECT, and cardiac 123I-MIBG scintigraphy) in predicting progression from mild cognitive impairment (MCI) to dementia were selected according to the Preferred Reporting Items for a Systematic Review and Meta-Analysis (PRISMA) method and evaluated with the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. Main eligibility criteria were as follows: (1) ≥ 50 subjects with MCI, (2) follow-up ≥ 3 years, (3) gold standard: progression to dementia or diagnosis on pathology, and (4) measures of prospective accuracy. Sensitivity (SE) and specificity (SP) in predicting progression to dementia, mainly to Alzheimer's dementia were 43-100% and 63-94% for FFDG-PET and 64-94% and 48-93% for amyloid-PET. Longitudinal studies were lacking for less common disorders (Dementia with Lewy bodies-DLB and Frontotemporal lobe degeneration-FTLD) and for tau-PET, DaT-SPECT, and 123I-MIBG scintigraphy. Therefore, the accuracy values from cross-sectional studies in a smaller sample of subjects (n > 20, also including mild dementia stage) were chosen as surrogate outcomes. DaT-SPECT showed 47-100% SE and 71-100% SP in differentiating Lewy body disease (LBD) from non-LBD conditions; tau-PET: 88% SE and 100% SP in differentiating DLB from Posterior Cortical Atrophy. I-MIBG scintigraphy differentiated LBD from non-LBD conditions with 47-100% SE and 71-100% SP. Molecular imaging has a moderate-to-good accuracy in predicting the progression of MCI to Alzheimer's dementia. Longitudinal studies are sparse in non-AD conditions, requiring additional efforts in these settings.