Abstract Background Primary central nervous system lymphoma (PCNSL) responds relatively quickly to chemotherapy or radiotherapy. However, determination of a complete response after treatment is often difficult because of extremely light residual contrast enhancement on magnetic resonance images due to the effects of microhemorrhages and scar tissue formation. These small enhancing lesions define an unconfirmed complete response. The aim of this study was to investigate the usefulness of carbon-11-labeled methionine (11C-Met) positron-emission tomography (PET) for determining the treatment response of PCNSL. Methods Data for 36 patients who were treated for PCNSL between 2011 and 2015 and underwent magnetic resonance imaging and 11C-Met PET were reviewed. Magnetic resonance imaging findings were classified as complete response, unconfirmed complete response, and tumor mass (a composite of partial response, stable disease and progressive disease). PET images were evaluated, standardized uptake values were quantified, and the tumor-to-normal tissue count ratio (TNR) was calculated. Receiver operating characteristic curves were generated to determine the optimal cutoff TNRs. Results The optimal TNRs for differentiating complete response and unconfirmed complete response from tumor mass were 1.83 (area under the curve, 0.951) and 1.80 (area under the curve, 0.932), respectively. The corresponding sensitivity and specificity values for the diagnosis of tumor mass were 82.4 and 100%, respectively, in the complete response group and 85.3 and 85%, respectively, in the unconfirmed complete response group. Conclusions A TNR of ≥1.80 can aid in the detection of active PCNSL using 11C-Met PET. Thus, 11C-Met-PET may be a useful tool for accurate evaluation of the treatment efficacy in PCNSL.