Infliximab is effective as a third-line therapeutic for severe sarcoidosis; however, long-term efficacy is unknown. The aim of this study was to assess the relapse rate after discontinuation of infliximab in sarcoidosis patients and predict relapse by analysis of the activity marker soluble interleukin (IL)-2 receptor (sIL-2R) and maximum standardised uptake value (SUVmax) of (18)F-fluorodeoxyglucose positron emission tomography (FDG PET). In this retrospective cohort study, the proportion of relapse was analysed using the Kaplan-Meier method and predicting factors were studied using Cox regression. 47 sarcoidosis patients who started infliximab therapy were included in the risk analysis. Kaplan-Meier analysis revealed a median time to relapse of 11.1 months and showed that 25% of the cohort relapsed within 4 months. Both mediastinal SUVmax ≥ 6.0 on FDG PET (hazard ratio 3.77, p<0.001) and serum sIL-2R ≥ 4000 pg · mL(-1) (hazard ratio 2.24, p=0.033) at start of therapy predicted relapse. In multivariate analysis, a mediastinal SUVmax ≥ 6.0 at initiation of therapy was an independent predictor of relapse (hazard ratio 4.33, p<0.001). The majority of patients that discontinued infliximab therapy relapsed. High serum sIL-2R and high SUVmax on FDG PET at initiation of therapy were significant predictors of relapse. These results suggest close monitoring of patients in this category when they discontinue infliximab treatment.