The low-quality black tea was extracted at 27 different conditions using a lab-scale supercritical fluid extraction system according to four factor, three level Box-Behnken design [pressure (150-450 bar), temperature (40-80 degrees C), modifier flow rate (0.5-1.0 ml/min), and ethanol concentration in aqueous solution (75-100%)] at constant CO2 flow rate (21/min). Response surface methodology was used in order to optimize the extraction conditions for obtaining minimum caffeine and maximum phenolic profiles of the decaffeinated black tea. The R-2 values for caffeine and phenolics were 99.5 and 96.6%, respectively. The lowest caffeine and the highest phenolics were obtained at following conditions [pressure (300 bar), temperature (53 degrees C), modifier flow rate (0.70 ml/min), and ethanol concentration (87.5%)] for 1 h. Using these conditions, the average loss of caffeine and phenolics in the decaffeinated tea were 99.8 and 3.3%, respectively. The present work suggests that optimum extraction conditions found can be applied for a pilot or large-scale production of decaffeinated black tea. (C) 2013 Elsevier B.V. All rights reserved.