To address the issue of low accuracy in the joint evaluation of smart meter errors and line loss rates, a joint evaluation model based on the improved sand cat swarm optimization (ISCSO) algorithm is proposed. First, the fitness function of the model is determined using a typical station topology structure and the law of energy conservation, and parameter ranges are determined based on area data. Next, the original sand cat swarm optimization algorithm is improved using a zoomed good point set, a Weibull optimal value-guided strategy, the dandelion optimizer, and an associative learning mutation strategy. The superiority of the improved algorithm is verified using benchmark test functions. Finally, the joint evaluation model for smart meter errors and line loss rates is established based on the fitness function and the improved algorithm. Case studies verify that the proposed method significantly improves the evaluation accuracy of smart meter errors and line loss rates compared to the dynamic line loss smart meter error evaluation model based on recursive least squares with a forgetting factor, and the constrained optimization model for joint evaluation of smart meter error and line loss rate.