Compared with an unselected susceptible population, a cotton bollworm, Helicoverpa armigera (Hübner), population selected for 22 generations with transgenic cotton leaves (modified Cry1A) in the laboratory developed 11.0-fold resistance to Cry1Ac (one single-protein product MVPII). Resistance to Bacillus thuringiensis Berliner subsp kurstaki (Btk) was selected for 22 generations with a 5.2-fold increase in LC50. The estimated realized heritabilities (h2) of resistance for transgenic-cotton- and Btk-selected populations were 0.1008 and 0.2341, respectively. This reflects the higher phenotypic variation in response to Cry1Ac in the transgenic-cotton-selected population. This variation may have been caused by differences in protein toxin levels expressed in different growth stages of the transgenic cotton. Because of the different slopes of the probit regression lines between Cry1Ac and Btk, the estimated realized h2 cannot be used visually to compare resistance development to Cry1Ac and Btk in H armigera. Thus, the response quotient (Q) of resistance was also estimated. The Q values of resistance for transgenic-cotton- and Btk-selected populations were 0.0763 and 0.0836, respectively. This showed that the rate of resistance development would be similar in both selection populations. This result indicates that the selection of resistance using transgenic cotton is different from that selected using the single toxin. Resistance risk to transgenic cotton and Btk in field populations was assessed assuming different pressures of selection by using the estimated h2. Assuming the h2 of resistance in a field population was half of the estimated h2, and the population received prolonged and uniform exposure to transgenic cotton or Btk causing >70% mortality in each generation, we predicted that resistance would increase 10-fold after <23 generations for Cry1Ac in transgenic cotton-selected-populations and after <21 generations for Btk in Btk-selected populations. Cross-resistance would be expected after <48 generations for Btk in transgenic-cotton-selected populations and after <21 generations for Cry1Ac in Btk-selected population. The results show that the potential to evolve resistance is similar in both transgenic-cotton- and Btk-selected populations, but that cross-resistance development to Btk is slower in transgenic-cotton-selected populations than cross-resistance development to Cry1Ac in Btk-selected populations.