Background and Aims We characterized the relationship between soil and leaf concentrations of selenium in a hyperaccumulator and a non-accumulator to test the hypothesis that hyperaccumulators take up selenium while non-accumulators exclude it. We examined plant performance metrics and the ability of selenium to protect against herbivory by spider mites. Methods Known hyperaccumulator and non-accumulator species within the genus Astragalus were grown under a range of selenium concentrations and measured for tissue selenium, extent of herbivory, and vigor. Results Both hyperaccumulators and non-accumulators either failed to meet even the lenient threshold or exceeded even the strict threshold for hyperaccumulation depending on soil concentration. Both had decreased herbivory with increasing leaf selenium, and both grew larger at higher levels of selenium regardless of herbivory, despite a negative impact of higher relative uptake. Conclusions The relationships between selenium dosage and tissue concentrations matched only some model predictions. Under these conditions, the bioconcentration factor was a better delimiter between species than the absolute tissue concentration. We provide evidence that despite the apparent cost of uptake, selenium can enhance the growth of hyperaccumulators even when herbivory is not a significant factor. We propose the term Belemental stimulation^for this phenomenon.