A fundamental principal of resistance management is that the more complex and potent a toxin mixture, the slower resistance will develop to the mixture in an insect population. Thus, to develop more complex and potent mosquitocidal bacteria, we genetically engineered Bacillus thuringiensis subsp. morrisoni PG-14 and Bacillus thuringiensis subsp. jegathesan, to synthesize, respectively, the binary (Bin) toxin of Bacillus sphaericus or a combination of Bin and the CytlA protein of Bacillus thuringiensis subsp. israelensis. Engineering these two larvicidal bacteria in general significantly improved their efficacy against fourth instars in comparison with their wild-type parental strains. For B. thuringiensis subsp. morrisoni PG-14, which naturally synthesizes Cyt1A, synthesis of Bin improved efficacy nine-fold (LC50 from 4.5 to 0.5 ng/ml) against Culex quinquefasciatus Say, although no improvement was observed (LC50 of 2 ng/ml for both strains) against Aedes aegypti L. For B. thuringiensis subsp. jegathesan, cosynthesis of Bin plus Cyt1A in combination with its normal complement of endotoxins improved efficacy 17-fold (LC50 from 34 to 2 ng/ml) against Cx. quinquefasciatus and 3.2-fold (LC50 from 68 to 21 ng/ml) against Ae. aegypti. Addition of Bin alone to B. thuringiensis subsp. jegathesan did not improve toxicity (LC50 from 68 to 65 ng/ml) against Ae. aegypti, indicating that CytlA synergized the activity of the endotoxins in this strain against Ae. aegypti. These results demonstrate that mosquitocidal efficacy of these strains and likely their resistance management properties can be improved significantly by increasing their toxin complexity and the amount of toxin they synthesize.