Measuring dopamine D₂ receptor occupancy levels using positron emission tomography (PET) is still widely unavailable. The objective of this study was to evaluate the accuracy of predicting D2 occupancy from the antipsychotic plasma level in patients with schizophrenia. Positron emission tomographic studies that measured plasma levels of antipsychotics and their corresponding D₂ occupancy levels were identified, using MEDLINE and EMBASE (last search: March 2010). Antipsychotics that were investigated in a total of 20 subjects or more were included. All data points for each antipsychotic were fit to a one-site binding model to estimate the total plasma concentration of each antipsychotic associated with a 50% occupancy (ED₅₀) of brain D₂ receptors. The mean prediction error and the root mean squared prediction error were used to measure the predictive performance of individual D₂ receptor occupancies from plasma drug levels derived from a one-site occupancy model using an ED₅₀ value calculated for each data point. A total of 34 treatment arms from 23 studies involving 281 subjects were included. The mean (95% confidence interval) prediction errors and root squared prediction errors were as low as 0.0 (-1.8 to 1.8) and 8.9 (7.6-10.2) for risperidone (n = 98); 0.0 (-3.5 to 3.5) and 15.1 (12.9-17.3) for clozapine (n = 75); -0.1 (-1.2 to 1.2), 0.0 (-1.9 to 1.9), and 4.6 (3.5-5.8) for olanzapine (n = 42); 0.1 (-3.4 to 3.5) and 9.9 (7.3-12.5) for haloperidol (n = 35); and -0.1 (-3.3 to 3.1) and 12.3 (8.8-15.7) for ziprasidone (n = 31), respectively. These findings suggest that D₂ occupancy of antipsychotics could be estimated with a high degree of accuracy using widely available plasma levels.