Although most Klebsiella pneumoniae liver abscesses have been reported to be monomicrobial, clinical outcomes have not been compared between antimicrobial therapy with and without anti-anaerobic coverage. A propensity score–matched cohort study was conducted using the 731 cases of K. pneumoniae liver abscess. Clinical outcomes were compared between a group discontinuing anti-anaerobic agents after K. pneumoniae identification and a group continuing. A total of 170 cases were matched at a 1:1 ratio using their propensity to discontinue anti-anaerobic agents. The McNemar's test showed no difference in mortality rates (1.8% for discontinuation versus 2.3% for continuation; P = 1.00) or relapse (1.8% versus 2.9%; P = 0.73) between groups. Early discontinuation of anti-anaerobic agents had no association with treatment failure by means of the generalized estimating equation model (odds ratio 0.48; P = 0.14) and the Kaplan–Meier method (P = 0.85) in matched groups. Early discontinuation of anti-anaerobic agents does not affect the clinical outcomes of patients with K. pneumoniae liver abscess.