High-resolution precipitation data, generated through dynamical or statistical downscaling of the global climate model output, provides critical information in regional climate change assessment and adaptation. While downscaling techniques aim at accurate precipitation construction at fine scales, most efforts on evaluation and calibration have focused on analysis at grid scale. Such analysis ignores the spatial connection of precipitation across model grids and the spatial structure of precipitation at the event scale, although hydrologic modeling and analysis using the downscaled precipitation require reasonable representation of the spatial structure of precipitation within watersheds. To fill this gap, we conducted an object-based precipitation evaluation of several dynamically and statistically downscaled products over the contiguous US (CONUS). Applying computer vision techniques, we identified precipitation events or objects from both decades-long downscaled datasets and observations. We quantified the dynamical/statistical model performance in reproducing various features of precipitation objects in the observations: total precipitation volume (Ptot), precipitation area (Atot), peak intensity (Ict) and spatial structure (sharpness). Results indicate a more consistent performance across all these features in dynamically downscaled products, while statistically downscaled products often only exhibit good statistics on Ict that is more relevant to grid-based analysis. Model performance also varies across different climate zones and seasons, as well as between extreme and non-extreme precipitation. Based on this comprehensive evaluation, we aim to provide overall guidance on the choice of downscaled products for specific regions, seasons, and precipitation object features. These findings and recommendations can inform precipitation-relevant modeling and analysis over CONUS, guide future developments to improve dynamical and statistical downscaling techniques, and provide actionable information for the climate impact assessment and adaptation communities.