This study aimed to investigate the relationship between cutting speed and surface roughness, the research method is used Meta-Analysis, this method must be through an extensive literature collection, finding consistent with the framework of this study. Therefore, this study integrated foreign EI, SCI journals 9 in 1998-2014 that coding analysis as samples. First, the various effects of statistical values convert into r, and calculate the success or failure of the estimates, the heterogeneity Q test, significant test ZC value, the true scale of maternal effect, 95% confidence interval, ability to explain variation statistic to verify and analyze the relationship of cutting speed and surface roughness. The results of this study have a high explanatory power of cutting speed and surface roughness, high homogeneity of the relevant research, no publication bias problem.