Vegetation canopy biophysical and biochemical parameters are key indicators for ecological monitoring, agricultural management and climate change studies. Accurate retrieval of these parameters is essential for understanding vegetation physiological status and ecosystem functioning. In this study, we used multi-output Gaussian process (MOGP) regression to simultaneously forecast multiple vegetation biochemical and biophysical parameters. The prediction model was trained on synthetic data generated by the PROSAIL model, and detailed spectral information from UAV hyperspectral imaging was used as input to estimate spatially resolved vegetation parameter values. MOGP regression is a promising solution due to its three main advantages: flexible kernel design, small sample efficiency and uncertainty quantification. The advantages of MOGP in prediction accuracy are demonstrated through validation against single-output Gaussian regression. This approach provides a new framework for hyperspectral vegetation parameter retrieval that strikes a balance between interpretability, computational efficiency and error-aware prediction.