Several sibling species of the leaf beetle Diorhabda elongata (Brullé) have been introduced into North America for the biocontrol of saltcedars (Tamarix spp.), but only one, D. carinulata (Desbrochers), has been extensively used in the field. The first open releases took place in 2001, and widespread defoliation occurred at sites infested by Tamarix ramosissima, T. chinensis, and their hybrid forms. The beetles failed, however, to establish at sites where other Tamarix species are targeted for control. In this study, we compared the preference and performance of three Diorhabda sibling species using adult choice and larval performance experiments on the two formally targeted Tamarix species: T. ramosissima and T. parviflora. In the adult choice experiment, a greater proportion of D. carinulata was found on T. ramosissima than on T. parviflora. For the other two sibling species, D. elongata (Brullé) and D. carinata (Faldermann), adults were found in similar proportions on the two host plants. In the larval performance experiment, larval growth and survival did not differ between Tamarix species for any Diorhabda type; however, D. carinata larval biomass was 35–50% greater than the other beetles regardless of host species. Based on the few adults of D. carinulata found on T. parviflora in the adult choice experiment, we do not recommend introducing this beetle at sites where T. parviflora is targeted for biological control. The species D. carinata seems especially promising for future release because its larvae gained substantially more biomass than the other beetles during the same time period on both Tamarix species.