The control and instrumentation (CI) circuits of engineering systems are equally critical as even a minor fault may leads to major shut down of plants or may leads to major accidents. Hence the (CI) circuits need to be designed considering both electrical and mechanical parameters. The cable materials, besides possessing good electrical properties should also have desired mechanical properties. Hence it becomes necessary to find suitable cable material that possesses required electrical and mechanical properties. The preparation of new cable material by suitably blending existing material provides better results. This paper presents a method of identifying the suitable blend ratio of Silicone Rubber (SiR) and Ethylene Propylene Diene Monomer (EPDM) using Generalized Regression Neural Network (GRNN). The five different compositions of SiR-EPDM blends (A-90/10; B-70/30; C-50/50; D-30/70; E 10/90) were prepared. The mechanical parameters like tensile strength (TS), elongation at break (EB), Hardness (H) and the electrical parameters like volume resistivity (VRY), surface resistivity (SRY), arc resistance time (ART), comparative tracking index (CTI), Breakdown Voltage (BDV), Dielectric Strength (DS), Dielectric Constant (DC) were measured as per ASTM/IEC standards. The GRNN model was trained using the measured data. The proposed GRNN model has been tested with new data sets using MATLAB-SIMULINK. The test result reveals that GRNN model can effectively identify the SiR-EPDM blend ratio in order to meet the required electro-mechanical parameters for any specific application. © 2019 IEEE.