Sustainable utilisation of the available grazing area acts to increase the profitability and productivity of livestock grazing and should consider animals and grass sward. The labour-intensive and time-consuming tasks of fencing, animal monitoring, and controlling forage availability on pasture are general obstacles to the wider implementation of grazing. Virtual fencing (VF) opens up new opportunities by reducing fencing labour and increasing flexibility. In this study, we investigated and validated the potential of animal monitoring via VF collars and combined this with Unmanned-Aerial-Vehicle (UAV) data to monitor animals and pasture continuously. 32 Fleckvieh heifers were equipped with VF collars (Nofence®, Batnfjordsøra, Norway) and divided equally into four groups. Each group was assigned to a 2-ha pasture divided into four paddocks, each grazed for three to four days. For all heifers, GPS data were logged by the VF collars and used to evaluate walking distance, lying time, and spatial pattern of movement. Lying time measured by the VF collars was validated via a confusion matrix by using observational data as a reference. Our results suggest that relevant information on basic animal behaviours can be extracted from VF collars. UAV campaigns were carried out pre-and post-grazing on each paddock. 3D reconstructions, which allowed the calculation of digital orthomosaics and digital surface models were created from UAV imagery. On that basis, the Red-Green-Blue Vegetation Index (RGBVI), the change of the RGBVI, which was used to determine herbage on offer changes pre- and post-grazing and the height change between surveys were calculated and analysed on a polygon grid (2.5 × 2.5 m²) per paddock. RGBVI detected herbage on offer changes were validated by using ground truthing data (R²=0.51). A random forest model to analyse animal active time (lying time excluded) spent per polygon grid cell as derived from RGBVI and height changes provided a mean R² of 0.43. In addition, we used generalised linear mixed effect models to evaluate the impact of day on pasture on cattle behaviour. Lying time decreased, walking distance increased, while the distribution of cattle became more even by day on pasture. These results appeared to reflect the forage decrease from pre- to post-grazing. We conclude that behavioural and UAV-based spatial pasture utilisation analyses in addition to the fencing functions of the VF system have the capacity for sustainable, fine-scale decision-support in grassland management on a polygon grid base (‘grid grazing’).