This thesis includes four papers based on three different prospective cohort studies on injury characteristics in men’s professional football. The same general methodology was used in all papers. Time-loss injuries and player individual exposure was registered for match and training separately. The general aim was to investigate potential internal and external risk factors for injury, with a focus on age, playing position, time in professional football, playing surface (artificial turf and natural grass), changes between surfaces and climate; and to evaluate the study methodology. Paper I was based on data collected between 2001 and 2010 from 26 top professional clubs in Europe; the UCL injury study. In total, 6140 injuries and 797 389 hours of exposure were registered. A decreased general injury rate was observed for newcomers compared with established players (hazard ratio (HR), 0.77; 95% CI 0.61-0.99). Using goalkeepers as a reference, all outfield playing positions had significantly higher age-adjusted injury rates. Using players aged ≤ 21 years as a reference, the overall adjusted injury rate increased with age, with a peak injury rate among players aged 29 to 30  years (HR, 1.44; 95% CI, 1.24-1.68). Paper II and Paper III are based on data collected during two consecutive seasons, 2010 and 2011, in the Swedish and Norwegian male first leagues. In total, 2186 injuries and 367 490 hours of football exposure were recorded. No statistically significant differences were found in acute injury rates on artificial turf (AT) compared with natural grass (NG) during match play (rate ratio, 0.98, 99% CI 0.79-1.22) or training (rate ratio 1.14, 99% CI 0.86-1.50) when analysing at the individual player level. However, when analysing at the club level, clubs with AT installed at their home arena had a significantly higher acute training injury rate (rate ratio 1.31, 99% CI 1.04-1.63) and overuse injury rate (rate ratio 1.38, 99% CI 1.14-1.65) compared to clubs with NG installed at their home venue. No association was found between frequent surface shifts and subsequent overuse injury risk (risk ratio 1.01, 95% CI 0.91-1.12). Analyses on the total cohort showed no difference in injury rates between clubs in the two climate zones (total rate ratio 1.01, 95% CI 0.92-1.10). Data included in Paper IV were collected during two consecutive seasons 2008 and 2009. During this period, two Norwegian elite football clubs were concurrently included in two research groups’ surveillance systems. The capture rate for match injuries was similar between the two audits, while the capture rate for training injuries was slightly higher with individual-based exposure recording. The inter-rater agreement in injury variable categorisations was in most aspects very high.