The wine industry needs to know the yield of each vine field precisely to optimize quality management and limit the costs of harvest operations. Yield estimation is usually based on random vine sampling. The resulting estimations are often not precise enough because of the high variability within vineyard fields. The aim of the work was to study the relevance of using NDVI-based sampling strategies to improve estimation of mean field yield. The study was conducted in 9 non-irrigated vine fields located in southern France. For each field, NDVI was derived from multi-spectral airborne images. The variables which define the yield: (berry weight at harvest (BWh), bunch number per vine (BuN) and berry number per bunch (BN)) were measured on a regular grid. This data-base allowed for five different sampling schemes to be tested. These sampling methods were mainly based on a stratification of NDVI values, they differed in the way as to whether NDVI was used as ancillary information to design a sampling strategy for BuN, BN, BW or for all yield variables together. Results showed a significant linear relationship between NDVI and BW, indicating the interest of using NDVI information to optimize sampling for this parameter. However this result is mitigated by the low incidence of BW in the yield variance (4 %) within the field. Other yield components, BuN and BN explain a higher percentage of yield variance (60% and 11 % respectively) but did not show any clear relationship with NDVI. A large difference was observed between fields, which justifies testing the optimized sampling methods on all of them and for all yield variables. On average, sampling methods based on NDVI systematically improved vine field yield estimates by at least 5-7 % compared to the random method. Depending on the fields, error improvement ranged from -2 % to 15 %. Based on these results, the practical recommendation is to consider a two-step sampling method where BuN is randomly sampled and BW is sampled according to the NDVI values.