In sports injury prevention field trials, data collectors are often club volunteers with considerable knowledge of the game but with limited detailed medical backgrounds or knowledge of formal scientific processes. The aim of this paper is to determine the agreement among trained primary data collectors (PDCs) with a sport science background and no prior involvement in data collection in a large randomised controlled trial. During the 'Preventing Australian Football Injury through eXercise' (PAFIX) project, player participation and injury data were collected by trained PDCs at training and games over the 2007 and 2008 playing seasons in 40 community level Australian football teams. PDC-collected data relating to player exposure and whether or not a player sustained an injury and subsequently left the field of play was compared to the same information from independent observers (IOs) who attended one randomly selected game for each of the 40 teams. There was 98% agreement between the PDC and the IO on game details (i.e., date, time, grade and score), 79% (ICC 0.9, 95% CI 0.85-0.95) agreement on the number of players per game and 76% (ICC 0.8, 95% CI 0.69-0.91) agreement on the number of injuries sustained in the games. There was 100% agreement on whether the player left the field for all injuries. This study found that exercise and sport science students are reliable data collectors in sports injury fieldwork studies.