Accurate estimation of biomass in grasslands is essential for understanding ecosystem health and productivity. Unmanned Aerial Vehicles (UAVs) have emerged as valuable tools for biomass estimation using canopy height models derived from high-resolution imagery. However, the impact of field disturbances, such as lodging and molehills, on the accuracy of biomass estimation using UAV-derived canopy height models remains underexplored. This study aimed to assess the relationship between UAV-derived canopy height and both reference canopy height measurements and dry biomass, accounting for different management systems and disturbance scenarios. UAV data were collected using a multispectral camera, and ground-based measurements were obtained for validation. The results revealed that UAV-derived canopy height models remained accurate in estimating vegetation height, even in the presence of disturbances. However, the relationship between UAV-derived canopy height and dry biomass was affected by disturbances, leading to overestimation or underestimation of biomass depending on disturbance type and severity. The impact of disturbances on biomass estimation varied across cutting systems. These findings highlight the potential of UAV-derived canopy height models for estimating vegetation structure, but also underscore the need for caution in relying solely on these models for accurate biomass estimation in heterogeneous grasslands. Future research should explore strategies to enhance biomass estimation accuracy by integrating additional data sources and accounting for field disturbances.