Abstract: The in-situ, real time, accurate and rapid detection based on mass spectrometry mostly focus on ionization technology and analytical method. However, the quantitative research of mass spectrometry performance factors that affect sensitivity, repeatability and constancy in the application is rather little. 240 groups of condition experiments for dimethyl methylphosphonate (DMMP) were conducted with gas chromatography-electron ionization-mass spectrometry (GC-EI-MS), and the analysis of variance (ANOVA) method was used to quantitative the influence extent of the inlet temperature of gas chromatography, the transmission line temperature between GC and MS, and the ion source temperature. The results of single factor ANOVA showed that the average F-ratio and p-value were 59.56 and 3.2×10^-4, 23.94 and 0.027, 68.12 and 0.022 (F_0.05(3,8)=4.07, F_0.05(4,10)=3.48 and the threshold of p-value is 0.05) for inlet temperature, transmission line temperature and ion source temperature, respectively. So through comparing the F-ratio and p-value, it could be included that inlet temperature was the most significant factor that affect the performance of MS, and ion source temperature was less, and transmission line temperature was least. The results of multi-factor ANOVA showed that F-ratio and p-value were 854.25 and 0, 112.79 and 0, 71.26 and 0 (F_0.05(3,160)=2.60, F_0.05(9,160)=1.88 and the threshold of p-value is 0.05) for inlet temperature, ion source temperature, the interaction between inlet and ion source temperature, respectively. So it could be further testified that the inlet temperature and ion source temperature were the significant factors. Experiments of linear quantify and repeatability based on the optimized factors was conducted with six different concentrations of DMMP. The results showed that linear quantify of GC-MS could cover 1.50 mg/L using the optimized factors, comparing with 10.50 mg/L of linear quantify without optimization. The average relative standard deviation (RSD) for DMMP with six different concentrations was about 20%, yet, RSD better than 5% could be achieved with 20.50 mg/L when using optimized factors. However, RSD without optimization was about 27%. The quantitative analysis of factors can provide data support for the enhancement of performance of mass spectrometry, especially for the sensitivity, linear quantify and repeatability.