The objective was to conduct a meta-analysis of studies of benzene exposure and non-Hodgkin lymphoma (NHL). A total of 8 cohort and 14 case-control studies were analyzed. Meta-analysis of any benzene exposure resulted in a summary relative risk estimate (SRRE) of 1.02 (95% CI: 0.94 to 1.12). The SRRE changed minimally when only data representing the highest level of benzene exposure were analyzed after an a priori data extraction protocol, using cumulative exposure as the optimum metric (SRRE = 1.08, 95% CI: 0.93 to 1.24). Meta-analysis of five studies that reported results for 60 or more ppm-years of cumulative exposure yielded an SRRE of 1.08 (95% CI: 0.36 to 3.24). Similarly, an SRRE of 1.04 (95% CI: 0.96 to 1.12) for each 25 ppm-year increment of benzene exposure was observed. The results from this meta-analysis are not supportive of an independent association between benzene exposure and NHL.