Optimization of medium components because of being multifactors is very difficult. In order to optimize the plant growth regulators (BA, IBA and GA3) for shoot proliferation of apple cv. ‘Abbasi’ a three-factors experiment was designed and performed in 20 points using Design Expert software according response surface method. A completely randomized design (CRD), based on eigth auxin combinations (IBA+NAA), also were performed for in vitro rooting of microcuttings. Proliferated shoot number was linearly and positively affected by the BA factor, while the response to the leaf number was linearly and negatively affected by BA. The effect of GA3 on proliferated shoot length response was observed in negatively linear model. The abnormal shoot percentage in a quadratic model was influenced by all three factors, BA, IBA and GA3; positively linear by BA, positively quadratic by IBA, and negatively quadratic by GA3. According to the results of this experiment, the optimum concentrations for BA, IBA and GA3 were 1.5, 0.1 and 0.02 mg L-1, respectively, which resulted proliferate shoots with a rate of 2.5 shoots per explant, with 0.9 centimeters in length, 4.3 leaves per shoot, and 4% of abnormal shoots. The percentage of shoot-tip necrosis was not affected by the evaluated factors. NAA was more effective than IBA for rooting of microcutings, and medium aupplelment with 1 mg L-1 IBA or 0.5 mg L-1 NAA can be used successfully for rooting of apple cv. ʽAbbasiʼ shoots in vitro.