building on previous studies on intellectual features and learners’ grammar learning, the present study aimed at investigating whether intelligence criterion had any impact on efl learners’ english grammar learning across two intelligence levels. in the current study, the participants were divided into two experimental and control groups by administration of raven i.q. test. this led to the formation of two intelligence levels, high average and superior, in two classes a and b. each group consisted of 28 participants totaling 56 i.q. levels. class a received grammar instruction explicitly, while the second group was instructed implicitly. the first testing material that was utilized in the present study was i.q. raven test in order to assign participants into intelligent proficiency levels. it is a non-verbal test with multiple choice measures of the reasoning which was held by the school’s counselor under the obliged schemata in iran’s i.q. testing program. language tests were taken from english course book grade 1 lesson eight (2011-2012) on active-passive voice to ensure the homogeneity of the participants and as a base to determine learners’ language capability. both pretest and posttest for the two classes (4 groups) were identical and consisted of totally 40 grammar items including 25 of active-passive voice. spss software and t-test with post hoc comparison illustrated that teachers’ instruction could not provide efficient impact on students’ english grammar learning when they are in different intellectual features. moreover, the gap seen in grammar learning could not be filled appropriately so that we claim all efl learners are able to achieve the same purpose of instruction, even if they differ one by one. among the two learning processes, learners tend more to a teacher-oriented class than the student-based class, even if they are in different intellectual capacity.