In patients with neurofibromatosis type 1, transformation of neurofibromas into a malignant peripheral nerve sheath tumor (MPNST) is a severe complication of the disease. Fluorine-18-fluorodeoxyglucose PET/computed tomography (PET/CT) is a viable option for detecting malignant tumors in neurofibromatosis type 1 patients. The aim of this review was to assess the diagnostic performance of the most frequently used parameters of PET/CT in detecting MPNST. An extensive computer search was performed using the Cochrane Library, Pubmed, and Medline/Embase databases. Two reviewers independently extracted data of relevant studies and assessed the methodological quality (QUADAS-2). The diagnostic performance of PET/CT parameters in individual studies was determined by calculating a diagnostic odds ratio (DOR) using the absolute numbers of true-positive, true-negative, false-positive, and false-negative test results. A total of eight studies were included, of which three evaluated the standardized uptake value as a diagnostic parameter, two assessed the tumor-to-liver (T/L) ratio, and three articles described both parameters. The cut-off values for maximum standardized uptake value (SUVmax) ranged from 3.2 to 4.5; for the T/L ratio, the cut-off values were between 1.0 and 4.3. The sensitivity and specificity ranged from 90 to 100% and from 80 to 100%, respectively (SUVmax). T/L ratios were associated with 92–100% sensitivity and 72–94% specificity. The corresponding DORs ranged from 57 to 145 (SUVmax) and 35 to 655 (T/L ratio). Both the SUV and the T/L ratio are associated with high sensitivity combined with acceptable specificity in detecting MPNST. There is a tendency toward higher DORs using the T/L ratio, but the number of studies is limited.