IntroductionThe overall prevalence of Klebsiella spp., a group of important zoonotic pathogens, in the global dairy herds and the risk of cross-species transmission between humans and dairy cows remain to be clarified. This systematic review aimed to determine the prevalence of Klebsiella spp. in milk samples from dairy cows with mastitis worldwide and to assess the factors influencing the prevalence of these strains.MethodsQualified studies published from 2007 to 2021 were retrieved from ScienceDirect, Web of Science, PubMed, WanFang Database, China National Knowledge Infrastructure (CNKI), and VIP Chinese Journal Database. Calculations of prevalence and their 95% confidence intervals (CIs) were performed for all the studies using the Freeman-Tukey double arcsine transformation (PFT).ResultsA total of 79,852 milk samples from 55 manuscripts were examined in this meta-analysis, and 2,478 samples were found to be positive for Klebsiella spp. The pooled prevalence estimates worldwide were 7.95% (95% CI: 6.07%–10.06%), with significant heterogeneity (I2 = 98.8%, p = 0). The sampling period of 2013–2020 had a higher (p < 0.05) Klebsiella-positive proportion of milk samples (12.16%, 95% CI: 8.08%–16.90%) than that of 2007–2012 (3.85%, 95% CI: 2.67%–5.21%), indicating that bovine mastitis caused by Klebsiella may become increasingly prevalent. The risk factors for the high prevalence of Klebsiella in milk samples mainly included: economic development level (developing countries; 11.76%, 95% CI: 8.25%–15.77%), mastitis type (CM; 11.99%, 95% CI: 8.62%–15.79%), and population density (>500 per sq km; 10.28%, 95% CI: 2.73%–21.58%). Additionally, a bivariate meta-regression analysis revealed that the multidrug-resistance (MDR) rate of the epidemic strains was also closely related to economic development level (R2 = 78.87%) and population density (R2 = 87.51%).DiscussionDue to the potential risk of cross-species transmission between humans and cows, the prevalence of mastitis milk-derived Klebsiella and its high MDR rate need to be monitored, especially in developing countries with high population densities.