OBJECTIVE: The aim of this study was to analyse the relationship between absolute and acute:chronic workload ratios and non-contact injury incidence in professional football players and to assess their predictive ability. METHODS: Elite football players (n = 130) from five teams competing in European domestic and confederation level competitions were followed during one full competitive season. Non-contact injuries were recorded and using session rate of perceived exertion (s-RPE) internal absolute workload and acute:chronic (A:C) workload ratios (4-weeks, 3-weeks, 2-weeks and week-to-week) were calculated using a rolling days method. RESULTS: The relative risk (RR) of non-contact injury was increased (RR = 1.59, CI95%: 1.18-2.15) when a cumulative 4-week absolute workload was greater than 10629 arbitrary units (AU) in comparison with a workload between 3745 and 10628 AU. When the 3-week absolute workload was more than 8319 AU versus between 2822 and 8318 AU injury risk was also increased (RR= 1.46, CI95% 1.08-1.98). Injury incidence was higher when the 4-week A:C ratio was <0.85 versus >0.85 (RR = 1.31, CI95%: 1.02-1.70) and with a 3-week A:C ratio >1.30 versus <1.30 (RR = 1.37, CI95%: 1.05-1.77). Importantly, none of the A:C workload combinations showed high sensitivity or specificity. CONCLUSIONS: In elite European footballers, using internal workload (sRPE) revealed that cumulative workloads over 3 and 4 weeks were associated with injury incidence. Additionally, A:C workloads, using combinations of 2, 3 and 4 weeks as the C workloads were also associated with increased injury risk. No A:C workload combination was appropriate to predict injury.