Purpose: The paper aims to critically examine overconfidence in numeracy among higher education (HE) graduates and its impact on their employability. The paper discusses the extent to which graduates, because of higher qualifications, overstate their numerical abilities. Design/methodology/approach: The paper is a review of the academic literature examining the theoretical significance of overconfidence in HE. The review subsequently draws on practice and policy reports that evidence graduates' overconfidence in numeracy and basic skills. Findings: The article shows a significant interaction between the level of qualification and overstatement of numerical abilities. The analysis found that graduates do not always have an important basic skill such as numeracy whose impact on work performance is significant. Practical implications: The findings are momentous for rethinking HE curricula, employee development in organisations and government skills strategy. The article advocates more inclusive and interpretive research for a greater understanding of the issues and offers useful data to policymakers and HE institutions in preparing graduates for work and decision-making. Further research in the field is required to enable the formulation of more authoritative conclusions. Originality/value: A critical contribution of this reflection is to have linked the evidence from the academic literature with employer surveys about graduate basic skills to draw the attention to a vital issue affecting national and organisational productivity, thus substantiating anecdotal evidence about graduate overconfidence. This reinforces the value of systematic literature review in research, as it provides an opportunity for more informed policy formulation as well as extending the body of research.