There is significant existing scholarship on the relationship between the English language proficiency (ELP) and academic performance (AP) of English Additional Language (EAL) learners at South African schools and universities. However, this relationship has not been explored in the Technical and Vocational Education and Training (TVET) sector in South Africa, particularly not among National Certificate (Vocational) (NC(V)) engineering students, where the failure rate is exceptionally high, and the graduation and certification rates disconcertingly low. Since English is the language of learning and teaching (LOLT) in the South African TVET sector, but the vast majority of TVET students speak English as an additional language, it is important to understand the implications of this for student progress and success. This article reports on the initial quantitative phase of a mixed-method study, conducted at one Eastern Cape TVET college to investigate this gap in existing knowledge. For the purpose of this article, ELP is defined based on Cummins's basic interpersonal communication skills (BICS) and cognitive academic language proficiency (CALP) distinction. The study found that NC(V) students' ELP is significantly associated with their throughput rate and their performance in each of their subjects. Poor ELP therefore acts as a barrier to students' academic progress in NC(V) engineering courses. Students' performance in the subject English First Additional Language (EFAL) in the initial year of study can be a predictor of academic progress and throughput. However, students' performance in the English component of the performance assessment of competency education (PACE) proficiency test administered at TVET colleges were found to not accurately predict their performance in English as a subject. This article can serve as a point of departure for further research about predictors of and barriers to ELP at TVET colleges and in trades-focused courses and for the development and implementation of suitable interventions.