A technical powder of Bacillus thuringiensis serovar. israelensis (B.t.i.) (VectoBac TP, 5,000 international toxic units [ITU]/mg), an aqueous suspension (VectoBac 12AS, 1,200 ITU/mg), and a granular formulation (VectoBac CG, 200 ITU/mg) were tested in the laboratory under different biotic and abiotic, conditions for efficacy against larvae of saltwater (Aedes taeniorhynchus) and freshwater (Culex nigripalpus) mosquitoes. Second-, 3rd-, and 4th-instar larvae of Cx. nigripalpus were 1.3-3-fold more susceptible to both VectoBac TP and VectoBac 12AS than were the respective larval instars of Ae. taeniorhynchus. For each species, 2nd-instar larvae were several-fold more susceptible to these B.t.i. preparations than were the 4th instars. In test cups, larvae under lower densities exposed to B.t.i. concentrations sustained 5-9-fold higher mortalities than larvae under high-density conditions. VectoBac TP and VectoBac 12AS stayed in suspension for up to 24 h with similar larvicidal efficacy, which was greater at 32-35 degrees C than at 15-20 degrees C. At 60 degrees C maintained for 24 h, VectoBac 12AS was adversely affected 2-3-fold in terms of potency, but VectoBac TP was not affected. Significant loss of potency of both VectoBac 12AS and VectoBac TP occurred when exposed to 35-37 degrees C under high light intensity (140,000-170,000 lux) for 6 h. Increasing salinity levels from 0 (fresh water) to 50% sea water caused gradual efficacy declines of VectoBac 12AS and VectoBac TP against Ae. taeniorhynchus larvae. VectoBac CG caused insignificant initial and residual (up to 8 days) larval mortalities of both mosquito species. This formulation did not release the active ingredient of B.t.i. in any significant mosquito larvicidal concentration in surface layers of water, and the formulation was more effective in shallower water. Storage of all 3 formulations under constant laboratory and variable field conditions for up to 8 months did not produce detectable potency loss of these products.