Objectives: To investigate predictors of a supportive training relationship and examination success using data from the first UK Public Health Training Audit. Study design: Secondary analysis of data from a national cross-sectional survey of UK public health trainees, conducted in April 2003. Methods: All UK public health trainees were sent a questionnaire including questions on training experience and examination performance. Possible predictive factors of a supportive trainer-trainee relationship and success at Part I and Part II of Membership of Faculty of Public Health examinations were first investigated in univariate analyses, and subsequently using logistic regression. Results: Two hundred and ninety responses were received (62% response). Supportive trainers gave feedback [odds ratio (OR)=11.2, 95% confidence interval (0) 3.7-34.0, P < 0.001], spent at least 1 h/week with their trainee (OR=5.2, 95% CI 1.7-15.4, P=0.003), held 3-monthly progress meetings (OR=2.9, 95% Cl 1.4-6.1, P=0.006), and encouraged wide training experience (OR=2.5, 95% Cl 1.2-5.4, P=0.016). Predictors of success in the Part I examination by the end of the second year of training were medical background (OR=4.4, 95% Cl 1.6-12.2, P=0.004) and perceived adequacy of examination support (OR=4.2, 95% CI 1.7-10.0, P=0.001). Predictors of success in the Part II examination by the end of the fourth year of training were access to examiners (OR=4.6, 95% CI 1.3-16.2) and wide experience (OR=4.1, 95% CI 1.1-14.5). Conclusion: Trainees feet supported when provided with adequate trainer contact time, feedback and wide experience. Part I examination success appears to be more likely amongst trainees with a medical background and those given examination.