A three-phase fluidized bed cell disruptor equipped with a multi-disk impeller was applied in the disruption of baker’s yeast cells for the release of alcohol dehydrogenase (ADH). The performance of cell disruption was assessed by the kinetics of disruption contributed by different operation parameters. The operating conditions of cell disruption, including rotational and linear speed of impeller, cell mass concentration, percentage volume of glass bead, duration of operation, and rate of gas bubbling, were optimized by design of experiment method. Based on the first-order model derived using path of steepest ascent, it was found that the rotational and linear speed of impeller, cell mass concentration, and percentage volume of glass bead were the significant process parameters. The optimal values of these parameters were then predicted using central composite design of experiments. As predicted with a second-order polynomial model, the maximum release of ADH was achieved under the operating conditions: agitator speed of 5000 rpm, 25% (ww/v) of cell mass concentration, 1500 mL of glass bead, 10 L/min of gas bubbling and operating duration of 5 min. The maximum release of ADH obtained from the experiment under the optimal conditions was 329.8 U/min/dw-cell, which was close to the value of 331.2 U/min/dw-cell predicted by the response surface model.