Precision agriculture relies on timely, detailed knowledge of crop status, even at plant level. In this context, high-resolution maps generated by UASs provide valuable information, e.g. through vegetation indices. This type of detailed dataset is nowadays available thanks to light and high performing sensors, like multispectral cameras, which can be airborne on UASs at low altitude over fields. Recently, 3D crop models, represented as point clouds or meshes, have enhanced remote sensing capabilities. These models can be derived from 3D sensors or structure-from-motion techniques using traditional imagery. This study investigates the potential of multispectral 3D point clouds in describing vine status and morphology. A case study was conducted in a vineyard in Agliano Terme, Italy (cv. Barbera). In August 2023, data were collected using a MicaSense Altum PT camera, airborne on a DJI Matrice 300 RTK. During each survey, the flight height (18 and 26 m above the ground) and/or the camera inclination (90°-nadiral and 60°) were adjusted. This led to a set of multispectral imagery blocks with different characteristics, which were used (singularly or in combination) to generate a set of vineyard 3D point clouds. The capability of properly assess the crop status was investigated by a set of parameters, at single plant scale, based on the spectral values distribution and morphology. The results showed the potential effectiveness of multispectral 3D point cloud in representing crops status, and provide prescriptions for good data acquisition procedures to enhance the quality of crop 3D model.