This study aims to assess human and machine translation of academic texts and critique the accuracy of these translations, the consistency of the translations in their use of specialized terminology, and the overall linguistic fluency of the translations. Although there is plenty of research relating to the translation of nonspecialized or literary texts, there is a pronounced shortage of studies attempting to evaluate the translation of academic texts, given the particular challenges of specialized vocabulary, syntax, and other structures. This study is intended to fill that particular gap. For this study, a corpus of 10 to 12 academic texts was evaluated using the MQM (Multidimensional Quality Metrics) model (a quality assessment model that captures both quantity and quality). The research findings show that although there is unparalleled speed and fluency in obtaining draft translations through the use of machine translation, there is typically no adequate use of contextualized or domain-specific terminology. This is in contrast to human translation, which is able to provide determinate accuracy, constancy, and appropriate use of specialized vocabulary, syntax, and structures. Other findings indicate that the amalgamation of human and machine translation is the optimal means of rendering academic texts, or at least the most effort-free means therein.