Dominance has been assessed in different ways in insecticideresistance studies, based on three phenotypic traits: the insecticideconcentration required to give a particular mortality(D_LC), mortality at a particular insecticidedose (D_ML), and fitness in treated areas(D_WT). We propose a general formula forestimating dominance on a scale of 0 to 1 (0 = completerecessivity and 1 = complete dominance).D_LC, D_ML, andD_WT are not directly related and their valuesdepend on genetic background and environmental conditions. We also showthat pest management strategies can have the consequence to increaseD_WT via the selection of dominance modifiers.Studies on resistance to Bacillus thuringiensis toxinsprovide the ultimate example of the complexity of the definition of theconcept of dominance. Almost all studies have focused on calculation ofD_LC, which provides little information about theefficiency of pest management programs. For instance, one assumption ofthe high dose/refuge strategy is that Bacillus thuringiensisresistance must be effectively recessive (i.e.,D_ML must be close to zero). However,D_WT, rather than D_ML, isrelevant to the resistance management strategy. Therefore, we stronglysuggest that the time has come to focus on fitness dominance levels inthe presence and absence of insecticide.