A nonpathogenic strain of Rhizobium vitis ARK-1 was tested as a biological control agent for grapevine crown gall. When grapevine roots were soaked in a cell suspension of strain ARK-1 before planting in the field, the number of plants with tumors was reduced. The results from seven field trials from 2009 to 2012 were combined in a meta-analysis. The integrated relative risk after treatment with ARK-1 was 0.15 (95% confidence interval: 0.07–0.29, P<0.001), indicating that the disease incidence was significantly reduced by ARK-1. In addition, the results from four field trials from 2007 to 2009 using R. vitis VAR03-1, a previously reported biological control agent for grapevine crown gall, were combined in a meta-analysis. The integrated relative risk after treatment with VAR03-1 was 0.24 (95% confidence interval: 0.11–0.53, P<0.001), indicating the superiority of ARK-1 in inhibiting grapevine crown gall over VAR03-1 under field conditions. ARK-1 did not cause necrosis on grapevine shoot explants. ARK-1 established populations on roots of grapevine tree rootstock and persisted inside roots for two years.