Research has pointed to the influence of the first language (L1) in the acquisition of the second (L2). In this study I investigate the interface between siSwati as an L1 and the acquisition of Academic English by students of the tertiary institutions of Swaziland. I examine five theoretical frameworks which are germane to L2 acquisition – error analysis, interdependency, transfer, interlanguage and fossilization. I discuss how these frameworks can help explain the low levels of proficiency in Academic English among learners in tertiary institution in Swaziland. In my research I employ qualitative research methods – questionnaires with both students and lecturers on initial and subsequent encounters with reading and writing both in the L1 (siSwati) and the L2 (English) – as well as quantitative research methods including statistical analyses of demographic and biographic data. In addition, in order to gauge the impact of the L1 on the L2 I analyse written texts of first and final year students at a number of tertiary institutions in Swaziland. Findings reveal that the students' L1 does, to some extent, interfere with their ability to properly acquire Academic English but cannot entirely explain the students' failure to acquire competency or near native proficiency in Academic English. Other militating factors include early educational environments which were not conducive to stimulating bilingualism, poor supply of text resources in both the L1 and the L2, the lack of a culture of reading in either the L1 or L2, the remoteness of English mother-tongue contexts, peripheral normativity practices in the institutions and indeed the emergence and development of a new variety of English in Swaziland. My own assessment criteria were critiqued during the course of this study and suggestions were made as to the validity of some of my assumptions about what constitutes "correct English". This insight should necessitate a new study on how English competency is assessed in Swaziland and to what it extent it is in line with contemporary views of what constitutes Standard English. It is hoped that the findings of this study will inform current debates on language teaching and assessment in tertiary institutions in Swaziland and also highlight areas of concern for academic programmes that focus on developing language and writing skills. Finally, I recommend that it is literacy in the L1 that needs to be addressed at the grass-roots route level in order for transfer to the L2 to occur successfully. Ultimately I conclude that efficient acquisition of Academic English can only be achieved when cognitive abilities have been properly developed in the L1.