IT security service protects their networked information asset and infrastructures using experts’ knowledge. To improve its service quality, by exploring quality factors and assessing the quality level, an approach based on a research model is required. Several research models have been proposed and used in various service area, but few cases of IT security service are found. In this work, a research model of service quality has been proposed, based on the existing research models such as SERVQUAL and E-S-QUAL. The factors which affect the quality of IT security service and the best quality measure including quantitative measures have been identified. By comparing them, the feasibility of using quantitative measures for quality measures has been addressed. For analysis, structural equation modeling and various statistical methods were used. The result shows that, among the three quality measures, satisfaction is the most significantly affected one by quality factors. The quantitative measures show positive correlation with other perceived quality measures. The analysis of the result provides several suggestions which can be used in measuring and improving the quality of IT security service. Emotional factors such as empathy are as important as the swiftness of service, which leads to a strategy of quality improvement based on customer satisfaction, not solely on technology only. And, in order to design reliable quantitative measures, standards such as rational target processing time should be first established. The work of this research will provide efficient tools to improve the quality of IT security service and help the quality research for other service areas.