Validity is considered to be of paramount importance in language testing, and therefore, remains the central concept to all designs and research activities in the field of testing and assessment. Arguably, all researches in language testing are in some senses about validity and the process of validation. In this regard, it is the intent of the present research to investigate the validity of the English language tests employed at the Higher Secondary level in Bangladesh. The research questions addressed concern finding out whether the tests are valid in terms of content and construct. The tests administered at this level are ‘achievement tests’, designed to measure the extent of learning in a prescribed content domain in accordance with explicitly stated objectives of a learning program. The first objective of the study is, therefore, to examine how far the course objectives are reflected in the contents of the existing tests. Secondly, the study makes an assessment of how well these tests measure the abilities they are intended to measure. The findings reveal a great mismatch between what the tests aim at testing and what they actually test. A wide gap is found between the curriculum goals and the existing test format. The study also finds that the Higher Secondary language tests are largely unable to measure the constructs they are based on. The key recommendations to increase the content and construct validity of these tests  include developing  test specifications and designing syllabus in accordance with  course objectives, using direct tests and authentic tasks, sampling widely and unpredictably, arranging  training programs for the language teachers, etc. Keywords: Validity, achievement test, test specifications, syllabus, direct test, authentic task