In the realm of electric vehicle design and development, critical vehicle parameters are estimated by coast down test. However, other drag factors can introduce errors to the estimation, including inertial and friction of rotating parts. Additionally, certain parameters are challenging to measure directly, such as drag coefficient, driveline losses and the frontal area of vehicles like go-karts. To address this challenge, a method of model-based parameter estimation is proposed in this research. An electric vehicle and powertrain model of a go-kart are developed, incorporating uncertain parameters like rolling resistance coefficient and air density. The target vehicle undergoes testing, with vehicle speed and motor torque recorded concurrently. Subsequently, the MATLAB/Simulink parameter estimation tool is applied to fine-tune the uncertain parameters, aligning the simulated vehicle speed and torque with the recorded test data. The optimization reduces speed and torque error between simulation values and test data by 24.76% and 15.58% respectively comparing to coast down model. Once optimized, the model is validated through simulation using another set of recorded drive cycle data reduces speed and torque error by 48.46% and 30.63% respectively comparing to coast down model. With a validated model boasting high confidence, design optimization tools are utilized to enhance the traction motor's pulley ratio. The optimized pulley ratio leads to a 33.80% improvement in energy efficiency, thereby enhancing the overall performance of the electric go-kart.