Rugby union is high-intensity contact based team sport that requires players to possess a diverse range of physical attributes. Numerous studies have highlighted differences in physical characteristics between playing levels and positions; however few studies have investigated longitudinal progressions of body composition and physical performance. Furthermore, the degree to which each component of fitness is relied upon in competition is unknown. Information gathered from studies of this nature will allow adolescent development programmes to be formulated to improve physical characteristics important for elite performance. Therefore the aims of this thesis were firstly, determine the differences and changes in physical characteristics in rugby union players; secondly, establish the relationship between physical characteristics and on-field performance; and thirdly, establish the effectiveness of an off-season physical conditioning programme in adolescent rugby union players. In the first study, a mixed modelling procedure was used to estimate the between-player differences and within-player changes in physical characteristics in 1161 rugby union players from 2004-2007. Differences between positions and playing level were consistent with those from previous research. However, 20-m sprint time was the only clear difference between Super Rugby (professional) and international players (backs and forwards, ~2.0%), indicating international players may be selected due to greater skill, tactical ability and experience. Small increases in strength (~6.5%) and small decreases in speed time (~1.6%) occurred as players moved from Super Rugby to either the provincial or mid-year international competitions. These changes may be a result of reduced training loads due to regular high-intensity matches and greater travel involved in the Super Rugby competition. Study two established the relationships between physical characteristics and on-field game behaviours. Sprint times (10 m, 20 m and 30 m) had small to moderate correlations (range r = -0.12 to -0.32) with game behaviours associated with high-intensity running (e.g. line breaks and tries scored). Repeated sprint ability and percent body fat had small to moderate correlations with activity rate (range r = -0.17 to -0.38). Although the low correlations indicate other factors may be associated with on-field performance, the relationships reflect the importance of optimal levels of speed, repeated sprint ability and body composition in order to effectively perform tasks within competition. The final study determined the efficacy of a 15-week supervised off-season conditioning programme compared to an identical unsupervised programme in 44 adolescent rugby union players. The short-term changes were assessed immediately post-conditioning, while the persistence of the effects were established after a 6-month unsupervised competition phase. Supervised training enhanced the gains in strength with small to large (range 9.1% to 50%) differences between the groups’ increases in one repetition maximum (1RM) following the conditioning programme. Strength declined in the supervised group during the unsupervised competition phase, resulting in only a small clear difference between the groups’ long-term change in box-squat 1RM (15.9%). Most other differences between the groups’ changes in body composition, speed and anaerobic and aerobic running performance as a result of training were trivial or unclear. There were unclear effects of age as a covariate on the changes in physical characteristics as a result of training. The greater gains made in the supervised training group during the conditioning phase may have been a result of greater adherence and overall training load. The lack of clear differences after the competition phase illustrates the importance of supervision if the aims of development programmes are to improve physical attributes. Speed, body composition and repeated sprint ability appear to be important physical characteristics in rugby union players due to superior performances by higher playing levels and their relationship with game behaviours. Substantial improvements in physical performance can be achieved in adolescent players after a structured, organised and supervised training programme. Therefore physical characteristics should be developed from an early age to ensure the player is physically ready for the demanding nature of professional rugby.