This study examined the relationship between the bowling workload of first-class fast bowlers and injury with the aim of identifying a "safe" fast bowling workload threshold. Twelve male fast bowlers (mean age 25 years) from an Australian state cricket squad were observed for the 1999--2000 cricket season. Workload was quantified by examining fixture scorecards and conducting surveillance at training sessions. Injury data were obtained from Cricket Australia's Injury Surveillance System. The seasonal incidence of injury was high with seven bowlers sustaining nine injuries. Whilst injured bowlers did not tend to bowl a greater number of deliveries on the day of injury, a significant increase in deliveries per session was observed in the 8-21 days prior to the date of injury (mean= 77) as compared with the rest of the season (mean= 60, p< 0.02). Bowlers with a weekly bowling workload above the mean of 203 deliveries were at an increased risk of injury (Risk Ratio (RR)= 6.0, 95% confidence interval (CI) 1.00-35.91). Those bowlers who bowled in five or more sessions in any 7-day period were also at an increased risk of injury (RR= 4.5, 95% CI 1.02 to 20.12). A consistent relationship between high bowling workload and injury was observed. The risk of injury was much higher for those bowlers with a sessional, weekly and monthly bowling workload above the group mean, especially when this high workload was consistent and sustained.