The present study compared two different types of English-language teaching approaches, the grammatical approach (GA) and the communicative approach (CA), by the cotwin control method. This study has two purposes: to study the effects of teaching approaches and to estimate genetic influences upon learning aptitudes. Seven pairs of identical twins (MZ) and 4 pairs of fraternal twins (DZ) participated in the experiment along with 68 other nontwin fifth graders. Each cotwin was assigned to the GA and CA respectively and received 20 hours of lessons over a 10-day period. The behavioral similarities between MZ cotwins were statistically and descriptively depicted. No major effect of either teaching approach was noted, but the genetic influence upon individual differences of learning achievement was obvious. Furthermore, an interesting interaction between the teaching approaches and intelligence was found, that is, that the GA capitalises on and CA compensates for intelligence. This interactional pattern could be interpreted as an example of genotype-environment interaction. The relationship between genetic factors and learning aptitudes is discussed.