Cloud and precipitation prediction poses considerable challenges in many numerical models. The global 13-km resolution System for High-resolution prediction on the Earth-to-Local Domains (SHiELD), a Unified Forecast System (UFS) prototype atmospheric model developed at the Geophysical Fluid Dynamics Laboratory (GFDL), is used to evaluate the prediction accuracy of clouds and precipitation. This research compares cloud and precipitation predictions from the 13-km SHiELD prediction system to observations. In particular, the geographic distribution, probability density function, diurnal cycle, error growth, and quantitative precipitation forecasts are carefully evaluated. In addition, to understand the extent of the prediction accuracy and its dependency on the model’s horizontal resolution, I compare the 13-km SHiELD with its 25-km and 6.5-km versions. The 13-km SHiELD shows excellent performance in ice water path, geographic mean precipitation, the global probability density function (PDF) for light to medium precipitation, PDF over land’s extreme precipitation, and peak precipitation time over the land. However, the complexities of cloud and precipitation prediction have resulted in noticeable biases in predicting the geographic distribution of precipitation, precipitation diurnal cycle, ice and liquid water path, and cloud fraction in SHiELD. We find that the SHiELD prediction system exhibits the potential for improving cloud and precipitation prediction using finer horizontal resolutions. Degradation in cloud fraction, ice water path, precipitation error, root mean square error, equitable threat score, area under the curve and fractions skill score occurs when the model’s resolution is reduced to 25 km. These comparisons help uncover and understand the biases in the SHiELD system, with the goal of proposing solutions for improving cloud and precipitation prediction.