A study was conducted to see whether the people of Burundi would learn English more efficiently if it were presented through the people's native language, Kirundi, rather than through French. Five components were determined for the experiment: lexicon, structure, conversation, pronunciation, and spelling. A survey of 23 advanced students in a university English department revealed divided opinions on the question. A survey of M.A. theses on the subject revealed that only one had advocated teaching English directly from Kirundi. This experiment attempts to test the thesis that the English tense system and the English language in general can be taught efficiently through the use of Kirundi for explanations. For about one month 375 students in their first year of secondary studies were taught some rudiments of English sounds, vocabulary, spelling, structure, or conversational skills. All had studied French for four years. Similar Kirundi and French groups were chosen in the same schools; ten short teaching programs were prepared and carried out by graduate psycholinguistics students. Higher performances were achieved in all groups taught directly from Kirundi, except in pronunciation. Lesson scripts and tests used in the experiment are appended. (AMH)