Objectives: Despite several consensus statements, different injury definitions are used in the literature. This study aimed to identify the impact of different injury definitions on the nature and incidence of complaints captured during a short-term running program for novice runners. Methods: 1696 participants completed weekly diaries on running exposure and musculoskeletal complaints during a 6-week running program. These data were used to compare six different injury definitions (presence of running-related pain, training-reduction, time-loss of.one day or one week). Injuries were registered under these different definitions. Consequently incidence and the nature of complaints were compared between definitions. Results: The different injury definitions resulted in incidences that varied between 7.5% and 58.0%, or 18.7 and 239.6 injuries per 1000 h of running. The median duration of injury complaints was 4-7 days for injuries registered under a 'day definition', while complaints registered under a 'week definition' lasted 20-22 days. For running-related pain injuries the median of the maximum amount of pain was 3.0. In training-reduction and time-loss injuries these median values were scored between 5.0 and 7.0. No significant differences in anatomical locations between injuries that were registered under a 'day definition' or a 'week definition' were found. Injuries registered under a time-loss definition were located relatively more often at the knee, while complaints at the pelvis/sacrum/buttock were captured more often under a running-related pain definition. Conclusions: Injury definitions largely impact injury incidence. Location of injury is also affected by choice of injury definition. This stressed the need for standardized injury registration methods. (C) 2015 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. Groningen, Netherlands.