The extraction conditions applied in destructive analysis of Hibiscus sabdariffa L. (Hs) matrices before the measurement of antioxidant compounds may reduce their activities during isolation and purification due to decomposition. Hence, near infrared spectroscopic system was applied to determine these compounds. The calibration models developed by using partial least squares regression (PLSR) with 48 Hs samples and 24 unknown samples were used to confirm the robustness of the developed model. The results of calibration models effectively allowed the determination of the concentrations of total phenol content (TPC), total flavonoid content (TFC), total antioxidant content (TAC), ferric reducing antioxidant power (FRAP), and 1,1-diphenyl-2-picrylhydrazyl (DPPH) with high correlation coefficients (0.85-0.91) and small standard error of cross-validation (RMSECV) ranged from 0.08 to 0.33. The prediction capacity (RPD) for TPC, TFC, TAC, FRAP, and DPPH ranged from 3.30 to 8.25, thus demonstrating that the near infrared spectroscopic (NIRS) equations developed were applicable to unknown samples. This study demonstrates that NIRS can be considered as a fast tool for the nondestructive determination of antioxidant compounds (phenolics and flavonoids) and antioxidant activity of the Hs.