Purpose: To investigate the effects of repeated-sprint training in hypoxia vs in normoxia on world-level male rugby union players’ repeated-sprint ability (RSA) during an international competition period. Methods: A total of 19 players belonging to an international rugby union senior male national team performed 4 sessions of cycling repeated sprints (consisting of 3 × eight 10-s sprints with 20 s passive recovery) either in normobaric hypoxia (RSH, 3000 m; n = 10) or in normoxia (RSN, 300 m; n = 9) over a 2-wk period. Before and after the training intervention, RSA was evaluated using a cycling repeated-sprint test (6 × 10-s maximal sprint and 20-s passive recovery) performed in normoxia. Results: Significant interaction effects (all P < .05, ${\eta }_{\mathrm{p}}^2>.37$) between condition and time were found for RSA-related parameters. Compared with Pre, maximal power significantly improved at Post in RSH (12.84 [0.83] vs 13.63 [1.03] W·kg⁻¹, P < .01, ${\eta }_{\mathrm{p}}^2=.15$) but not in RSN (13.17 [0.89] vs 13.00 [1.01] W·kg⁻¹, P = .45, ${\eta }_{\mathrm{p}}^2=.01$). Mean power was also significantly enhanced from Pre to Post in RSH (11.15 [0.58] vs 11.86 [0.63] W·kg⁻¹, P < .001, ${\eta }_{\mathrm{p}}^2=.26$), whereas it remained unchanged in RSN (11.54 [0.61] vs 11.75 [0.65] W·kg⁻¹, P = .23, ${\eta }_{\mathrm{p}}^2=.03$). Conclusion: As few as 4 dedicated specific RSH sessions were beneficial to enhance repeated power production in world-level rugby union players. Although the improvement from RSA to game behavior remains unclear, this finding appears to be of practical relevance as only a short preparation window is available prior to international rugby union games.