Unauthorized Accesses of Computer Application and Network Are Biggest Issue In the Field of Computer Security. Due to the Popularity of the Computer System and Computer Applications, New Kinds of Attacks Are Emerging Rapidly. the Intrusion Detection Systems (Ids) Detects This Type of Intrusion Activity. This Paper Is Design and Development of Ids Using Mobile Agent to Enhance the Efficiency As Compared to Previous Ids Based on Mobile Agents. Furthermore This Paper Is Also the Deep Study of Significant Parameters Called Features of Nslkdd Data Set and Its Values. on the Basis of Deep Analysis of 41 Feature of Nslkdd Dataset, 13 Significant Features and Its Threshold Values Identified By Proposed Ids. Total Five Rules For the Identification of Normal and Abnormal (Dos, U2r, Probe, R2l) Packet Is Designed Which Is Based on 13 Significant Features. Our Deep Analysis of Significant Parameters and Experiments Clearly Shows That Proposed Ids Is More Effective and Accurate Which Is Also Reducing Response Time and False Alarms Rate.