Objective To evaluated the efficacy and acceptability of antidepressants for the treatment of panic disorder. Methods We searched Medline， Embase， CBM， CNKI in both English and Chinese databases， for the head-to-head randomized controlled trial of antidepressant treatment for panic disorder. Quality assessment and bias evaluation were conducted with the Cochrane Collaboration's RevMan 5.1.0 software. The network Meta-analysis were conducted with the Stata 14.2 software. Results A total of 25 trials met the inclusion criteria， involving 10 medications： mirtazapine， paroxetine， clomipramine， citalopram， fluoxetine， venlafaxine， escitalopram， sertraline， reboxetine， amfebutamone. The result of analysis showed that the drugs with good effectiveness were escitalopram， fluoxetine， venlafaxine and paroxetine， but there were no statistical significance in pairwise comparison of drugs （P＞ 0.05）. The more prominent antidepressants for the effect of score reduction were fluoxetine， mirtazapine， escitalopram， amfebutamone. Compared with clomipramine， fluoxetine ［MD=4.27 （1.12，7.43）］， escitalopram ［MD=3.84 （1.49，6.19）］， mirtazapine ［MD=3.42 （0.83，6.01）］， paroxetine ［MD=2.81 （0.82，4.80）］ and venlafaxine ［MD=2.06 （0.11，4.02）］ had obvious advantages in the effect of score reduction， which could reduce the score of Hamilton Anxiety Rating Scale （HAMA） significantly. There were no statistical significance in other pairwise comparisons of these drugs （P＞ 0.05）. The drug with good tolerance was paroxetine， followed by venlafaxine and escitalopram， but there was no significant difference in pairwise comparison of these drugs （P＞ 0.05）. Conclusions As the first-line drugs for the treatment of panic disorder， paroxetine， venlafaxine and escitalopram are effective drugs with both efficacy and safety.