OBJECTIVE: Data from Q fever pre-vaccination screening were analysed to determine the level of agreement between the two tests of immunity and between disease or vaccination history and immunity, trends in proportion of participants immune to Q fever, and the annual risk of infection. METHOD: Data from nearly 10,000 screening episodes between July 1988 and June 2001 on Victorian workers at high risk were assessed. RESULTS: Most participants were male (86%) and employed in Victorian abattoirs (81%). Agreement between results of tests for immunity was ‘fair’ (kappa=0.52). Self-reported history of vaccination or infection was a poor predictor of immunity. The proportion of positive blood and skin tests increased with years of exposure to animals/meat, but decreased over the 13-year period with shorter exposures. Hence the percentage requiring vaccination increased from 50% in 1998 to 90% in 2001. The average annual risk of infection among abattoir workers was 45.0 per 1,000 (95% CI 42.3-47.6), and 62.6 per 1,000 (95% CI 57.5-67.7) over the first 10 years of exposure. CONCLUSIONS: This is the first Australian study to estimate the annual risk of Q fever infection in abattoir workers. The study confirmed previous findings of poor agreement between screening tests and predictive value of history of vaccination/ exposure. Up to 90% of new entrants in high-risk workplaces will be susceptible to Q fever and require vaccination. IMPLICATIONS: Systematic post-marketing surveillance is needed to monitor adverse events to the vaccine, duration of protection and possible reasons for vaccine failures.