Background Numeracy in the general population is low: approximately 50% of adults of working age have numeracy skills at the level expected of an average schoolchild at age 11. There is increasing evidence that many students entering university across the globe have numeracy levels significantly below those required for their chosen courses. Clinician Numeracy (CN) is the ability to understand and use numerical information and quantitative data of all kinds in delivering safe treatment to patients. Research demonstrates that many medical students and doctors worldwide have low clinician numeracy, with implications for safe patient care. Methods My research includes the evaluation of various assessments of CN, followed by a comprehensive review of one assessment, the Medical Interpretation and Numeracy Test. I have conducted two randomised controlled trials: the first to assess whether calculators influence CN test results, and the second to investigate whether answer format (multiple choice or constructed response) influences CN test results. Finally, I investigated and classified the types of error being made by test participants. Results Following analysis, the Medical Interpretation and Numeracy Test appeared to be the most appropriate measure of CN for medical students and doctors. The Medical Interpretation and Numeracy Test was subjected to an emendation process, and psychometric analysis indicated that it is a reliable and valid test of CN. The revised test was used for my research. Approval for this research was granted by the University of Manchester Research Ethics Committee. All research was carried out on third year medical students in a single institution in England. The first randomised controlled trial showed that calculators did not affect test scores. The second showed that the multiple choice (single best answer) format of the test was equivalent to the constructed response (very short answer) format. The exploration of error demonstrated that most errors related to a failure to set the problem up correctly; furthermore, the type of errors being made were basic mistakes, similar to those made by nursing and biomedical science students. Conclusion This research confirms that the level of CN in medical students is variable, and that some have low numeracy. This is consistent with national and global data relating to low numeracy in the general population, in schoolchildren, in university students, and in healthcare professionals. Medical schools must ensure their graduates are competent to provide safe patient care; graduates have a responsibility to identify and remediate areas of weakness that may affect their practice. It is time for Clinician Numeracy to be included in medical curricula.