PhD (English) / Department of English / This research assesses how the choice of a particular medium of instruction supports Primary Class Three (P.3) pupils’ academic performance in literacy. The need for this research has been occasioned by the inconsistencies that exist in Ghana’s language-in-education policy. The current language-in-education policy allows for only the dominant language of the community, in which a school is situated, to be used as medium of instruction, even when pupils speak different languages in one P.3 classroom. The research adopts a mixed methodology approach and uses purposive sampling technique to select a total sample of 317 participants. This number comprises pupils, teachers, parents and other stakeholders in education from 8 schools within 3 municipalities in the Greater Accra region of Ghana. Five (5) of the schools use English medium of instruction and three (3) schools use Ghanaian language or mother-tongue. Data collection tools for the research are interviews, participant’s observations, video recordings and diagnostic tests. The main finding, based on the test results of the research, establishes that when mother-tongue and English are compared as mediums of instruction from Kindergarten One (KG1) up to P. 3, there is no significant difference pupils’ academic results in literacy. In view of this, the research advocates for further debate on other factors that support pupils academic performance in literacy, in addition to mother-tongue and English medium of instruction in lower primary schools.