The Plackett-Burman multifactorial design was employed to screen the important malting parameters for superoxide dismutase (SOD) in final malt of Ganpi-3. The eight factors screened for SOD were steeping temperature, steeping time, peroxide hydrogen concentration in steeping water, germination temperature, germination time, withering temperature, drying temperature and kilning temperature. Variance analysis showed that steeping time, germination temperature and kilning temperature were significant for SOD activity. Box-Behnken experimental design was further used to optimize the levels of the above three factors. By response surface methodology and canonical analysis, the optimal malting factors for higher SOD activity in final malt were: steeping time 42.2 h, germination temperature 16.9°C and kilning temperature 82.9°C. Under these conditions, the model predicted a SOD activity of 2234 U/g of dry weight malt. Verification of the optimization showed that a SOD activity of 2220 U/g was observed under optimal conditions. It showed that the experimental data could be reliably predicted by the polynomial model. Besides Ganpi-3, three other barley varieties including Ganpi-4, Ken-2 and Hamelin were malted under optimal and common conditions under laboratory conditions. To some extent, SOD activities were higher in malts from the optimal malting process than those from the common malting process. Especially, SOD activities in Ganpi-3 and Hamelin increased by 18.8% and 15.3%, respectively. Furthermore, twenty-nine samples of malts, including eleven imported malts and eighteen domestic malts, were used. Relationships between SOD activity in malt and the reducing power of wort were examined. There was significant correlation between SOD activity and the reducing power of wort (R 2 = 0.8069).