Abstract The economic success of many drilling operations depends on the availability and reliability of real-time information about the drilling process. Mud pulse telemetry is currently the most common method of transmitting measurement-while-drilling (MWD) and logging-while-drilling (LWD) data. Advances in downhole sensing for drilling optimization and formation evaluation are placing heavy demands on telemetry systems to provide fast and reliable data rates from greater depths. However, solid particle erosion poses a significant problem for telemetry tools, where solid particle (such as sand) impingement could damage the tool string and shorten the service life of the tools. Therefore, a comprehensive investigation on erosion of mud pulse telemetry tools consisting of numerical simulation and field tests is often required to optimize the tool design. In the field, many factors can influence telemetry tool erosion such as material properties, sand size, geometry, flow velocity, operating pressure, and turbulence. These factors interact with each other, making the experimental study of all influencing parameters very challenging and time-consuming. In this work, computational fluid dynamics (CFD) simulations were used to study the effect of several parameters on the erosion rate, even in complex geometries where setting up an experimental study is difficult. The erosion rate was determined using the widely used Oka erosion model. Parameter studies were then performed to find the influence of flow rate and sand concentration on the erosion rate. Simulation was also performed to support the deployment of new engineered materials. For model validation, simulation results were compared with erosion patterns from field tests, showing good agreement between field observations and simulation results. Based on findings from the parameter studies, a formula of key performance indication (KPI) parameter was developed to evaluate the erosion performance of the mud pulse telemetry tools deployed in the field. After completing the field experiments, 3D laser scans of the deployed tools with different materials were performed. In addition, KPI values were calculated based on the scanning results to evaluate the actual erosion performance. Evaluation revealed that the new engineered alloy was eight times more erosion-resistant than stainless steel, which was consistent with the CFD simulation results. The results of this study indicate that CFD simulation provided an alternate way to predict solid particle erosion on logging tools in downhole environments. By using the high-fidelity erosion model, the tool erosion rate could be accurately predicted. Based on this conclusion, the erosion risk can be mitigated by providing guidance on repair and maintenance intervals and planning the drilling process to avoid premature tool failures. This approach will eventually improve the reliability and safety of downhole tool and reduce non-productive time (NPT) and costs.