A postal survey was conducted regarding statistical techniques, research methods and software used in the workplace by 913 graduates with PhD and Masters degrees in the biological sciences, psychology, business, economics, and statistics. The study identified gaps between topics and techniques learned at university and those used in the workplace, and points to deficiencies in statistical preparation for employment. Courses requested include multivariate statistics, generalized linear models, research design and power analysis taught with minimal emphasis on probability and mathematics. Recommendations are presented, such as expanding statistical service courses to eliminate gaps, the development of intensive workshops for postgraduate students and for workplace retraining, or involving staff from other departments to provide context for statistics teaching. (Contains 7 tables.)