ObjectivesTo determine the incidence of illness, and identify the relationship between sleep, training load and illness in nationally competitive Australian football athletes. Second, to assess multivariate effect between training load and/or sleep variables. DesignCohort study. MethodsRetrospective analyses of prospectively collected cohort data were conducted on forty-four male athletes over a 46-week season. The primary outcome was illness incidence, recorded daily by medical doctors. Independent variables were acute, chronic and acute:chronic ratios of: sleep quality, sleep quantity, internal training load and external training load defined as: total running distance, high speed running distance and sprint distance. Generalised estimating equations using Poisson (count) models were fit to examine both univariate and multivariate associations between independent variables and illness incidence. Results67 incidences of illness were recorded, with an incidence rate of 11 illnesses per 1000 running hours. Univariate analysis showed acute and chronic sleep hours and quality, as well as acute sprint and total running distance to be significantly associated with illness. Multivariate analysis identified that only acute sleep quantity was significantly, negatively associated with illness incidence (OR 0.49, CI 0.25–0.94) once all univariate significant variables were controlled for. There was no relationship between external training load and illness when sleep metrics were controlled for. ConclusionsIn a cohort of Australian football athletes, whose load was well monitored, reduced sleep quantity was associated with increased incidence of illness within the next 7 days. Monitoring sleep parameters may assist in identifying individuals at risk of illness.