In this work, diagnosis of boron (B) nutritional status based on leaf B concentrations was compared for the most common leaf-sampling times for olive trees, January and July. For this purpose, field experiments were conducted over 4 years (2003–2006) in two rainfed olive groves located in Mirandela and Bragança, northeastern Portugal. Leaf samples were collected in January and July and analyzed for B by standard procedures. Fruit harvest occurred in December of each year. The crops followed typical alternate fruiting cycles. During the 4 years of the study, mean olive yields in the Bragança orchard fluctuated, yielding 3.6, 28.1, 5.5, and 22.7 kg tree−1. Yield variation per individual tree was also great. In the Bragança orchard and in the 2004 harvest, yields ranged from 1.2 to 52.7 kg tree−1. Leaf B concentrations also varied greatly between individual trees. In the Bragança orchard in the July sampling of 2004, values for individual trees varied from 12.2 and 23.7 mg B kg−1. From a total of 16 scatterplots generated from the relationship between leaf B concentrations and olive yields, 10 significant linear relationships were established; six of them were related to July sampling dates and four were related to January dates. The number of significant linear relationships established between leaf B concentration and olive yield was used as a criterion of the accuracy of the B nutritional diagnosis, because this represented the lowest experimental variability. By using this criterion, the July sampling date proved as better for B nutritional diagnosis, although the difference from January sampling date was not sufficient to disregard this. Leaf B concentrations were consistently greater in July than in January. Averaged across the 4 years of the study in both orchards, the difference was 4.3 mg B kg−1. This difference should be taken into account in the interpretation of leaf analysis results when B levels are close to the deficient critical concentration.