The objective of this study was to compare incidences and patterns of injury for female and male elite teams when playing football on artificial turf and grass. Twenty teams (15 male, 5 female) playing home matches on third-generation artificial turf were followed prospectively; their injury risk when playing on artificial turf pitches was compared with the risk when playing on grass. Individual exposure, injuries (time loss) and injury severity were recorded by the team medical staff. In total, 2105 injuries were recorded during 246 000 h of exposure to football. Seventy-one percent of the injuries were traumatic and 29% overuse injuries. There were no significant differences in the nature of overuse injuries recorded on artificial turf and grass for either men or women. The incidence (injuries/1000 player-hours) of acute (traumatic) injuries did not differ significantly between artificial turf and grass, for men (match 22.4 v 21.7; RR 1.0 (95% CI 0.9–1.2); training 3.5 v 3.5; RR 1.0 (0.8–1.2)) or women [match 14.9 v 12.5; RR 1.2 (0.8–1.8); training 2.9 v 2.8; RR 1.0 (0.6–1.7)]. During matches, men were less likely to sustain a quadriceps strain (P=0.031) and more likely to sustain an ankle sprain (P=0.040) on artificial turf.