Determining the relationship between soil B and crop B content can help predict when crops will respond to B fertilizer and when B toxicity may be expected. Such a relationship can then be used to make fertilizer recommendations or to flag conditions of potential B toxicity. Soil samples were obtained from 65 sites located in the Broadview Water District in the San Joaquin Valley of California. A diverse set of extractants was evaluated including: hot water-soluble, 1:1 soil:distilled water and 1:2 soil:distilled water, ammonium acetate, calcium chloride-mannitol, and DTPA-sorbitol extracts. Soil extract B values were correlated significantly with various B reactive soil constituents, including aluminum and iron oxide, clay, organic matter, and calcium carbonate content. The 1:1 water extract B was highly significantly correlated (99% level) with other measures of extractable B used in the study. Extractants were compared on soil samples collected from six depths at 65 field sites in the San Joaquin Valley of California that were cropped to alfalfa, melons, and cotton. Boron concentrations of whole plants and composites of 10 leaves were determined. Plant sampling occurred at the time of soil sampling for the alfalfa. Cotton and melons were sampled at flowering and prior to fruit set, the recommended growth stages, respectively, for tissue sampling, and 6 weeks thereafter. Five weeks later the cotton was sampled a third time. Significant correlations (95% level) between extractable soil B and plant B were found for melons and cotton but not for alfalfa. Correlation coefficients for the ammonium acetate, DTPA-sorbitol, and 1:1 water extract were not statistically significantly different (95% level). Although significant correlations (95% level) were obtained, the equations provided relatively poor predictive capability. These results illustrate the difficulty of predicting plant B content based on soil B analyses from a single soil sampling.