The diagnosis of NCC depends on neuroimaging and serological confirmation. While antibody detection by enzyme-linked immunoelectrotransfer blot (EITB) fails to predict viable NCC, EITB banding patterns provide information about the host's infection course. Adding antigen ELISA results on EITB banding patterns may improve their ability to predict or rule out of viable NCC. We assessed whether combining EITB banding patterns with Ag-ELISA improves discrimination of viable infection in imaging-confirmed parenchymal NCC. EITB banding patterns were grouped into classes using latent class analysis. True-positive and false-negative Ag-ELISA results in each class were compared using Fisher's exact test. Four classes were identified: 1 (EITB-negative or positive to GP50 alone [GP50 antigen family]), 2 (positive to GP42-39 and GP24 [T24/42 family], with or without GP50), 3 and 4 (positive to GP50, GP42-39 and GP24, and reacting to bands in the 8-kDa family). Most cases in classes 3 and 4 had viable NCC (82% and 88%) compared to classes 2 and 1 (53% and 5%). Adding positive Ag-ELISA results to class 2 predicted all viable NCC cases (22/22 [100%]), whereas 11/40 patients (27.5%) Ag-ELISA negative had viable NCC (P < 0.001). Only 1/4 patients (25%) Ag-ELISA positive in class 1 had viable NCC, whereas 1/36 patients (2.8%) Ag-ELISA negative had viable NCC (P = 0.192). In classes 3 and 4, adding Ag-ELISA was not contributory. Combining Ag-ELISA with EITB banding patterns improves discrimination of viable from non-viable NCC, particularly for class-2 responses. Together, these complement neuroimaging more appropriately for the diagnosis of viable NCC.