Clomazone has been successfully used for weed control in rice, but crop injury is a potential problem on light-textured soils. Experiments were conducted to determine the effect of soil characteristics and water potential on plant-available clomazone and rice injury. A centrifugal double-tube technique was used to determine plant-available concentration in soil solution (ACSS), total amount available in soil solution (TASS), and Kd values for clomazone on four soils at four water potentials. A rice bioassay was conducted parallel to the plant-available study to correlate biological availability to ACSS, TASS, and Kd. TASS was significantly different in all soils. The order of increasing TASS for the soils studied was Morey < Edna < Nada < Crowley, which correlated well with soil characteristics. The order of increasing TASS after equilibrium was − 90 < − 75 < − 33 < 0 kPa. TASS values at 0 kPa were greater than two times the TASS values at − 90 kPa. It appears that severe rice injury from clomazone on these soils could occur if TASS > 110 ng g−1 and Kd < 1.1 ml g−1. We propose that the double-tube technique provides a more accurate estimate of available herbicide because the solution–soil ratios are < 0.33:1 and would be more representative of a plant root–herbicide relationship. This technique or some variation possibly could be further developed such that clomazone rates could be more clearly defined particularly on lighter-textured soils. TASS may be a better predictor of plant-available herbicide than ACSS when evaluating moderately to highly water-soluble herbicides in a nonsaturated soil environment.