BACKGROUND: Mycobacterium avium complex (MAC) infection, requiring prolonged treatment with an antibiotic combination, is an emerging public health concern. METHODS: This meta-analysis compared the efficacy of 2-drug regimens with that of 3-drug regimens in bacteriological responses, acquired macrolide resistance (AMR), and mortality among adult patients with disseminated MAC infection (DMAC) and MAC pulmonary disease (MAC-PD). RESULTS: Seven randomized controlled trials (RCTs) and 3 non-RCT studies, encompassing 1369 patients, were included. Treating DMAC with 2-drug regimens was associated with comparable bacteriological responses (odds ratio (OR) = 0.76, 95 % confidence interval (CI) = 0.48-1.18, P = .22) and mortality (OR = 1.29, 95 % CI = 0.59-2.83, P = .52), but had a higher risk of AMR (OR = 2.99, 95 % CI = 1.10-8.13, P = .03). For MAC-PD, 2-drug regimens were noninferior to 3-drug regimens in bacteriological responses (OR = 0.82, 95 % CI = 0.53-1.25, P = .35) and AMR (risk differences (RD) = 0.01, -0.02 to 0.05, P = .39), with no observed mortalities. Although not statistically significant compared to the 3-drug regimen, the macrolide-rifamycin regimen resulted in attenuated bacteriological responses (OR = 0.51, 95 % CI = 0.14-1.90, P = .32). However, the proportion of patients with bacteriological response (OR = 1.54, 95 % CI = 0.78-2.93, P = .23) and AMR risk (RD = 0.01, -0.02 to 0.04, P = .50) was not different between those under the macrolide-ethambutol regimen and those under 3-drug regimens. CONCLUSIONS: The macrolide-ethambutol 2-drug regimen may be a viable alternative option for treating MAC-PD, whereas DMAC can be preferably managed with a 3-drug regimen.