Abstract The physiological effects of strategies for a rapid loss of body mass immediately before weighing-in for competition in weight-governed sports are unclear. This study examined the effects of a 3%–4% loss in body mass on a boxing-related task. Seven novice amateur boxers completed three 3 min rounds of simulated boxing on a prototype boxing ergometer in an euhydrated state (E-trial) and after exercise-induced thermal dehydration (D-trial). All subjects lost body mass following dehydration–mean body mass fell 3.8 (SD ± 0.3)% [77.3 (SD ± 11.3) to 74.4 (SD ± 10.7) kg, P < 0.001] – but changes in plasma volume (PV) were inconsistent. Four subjects suffered reductions in PV between 15% and 30%, one subject maintained his E-trial value and two recorded an increase. The D-trial mean PV value was 8.0 (SD ± 17.2)% lower but this fall was not statistically significant (P > 0.05). Analysis of D-trial boxing performance showed one subject maintained his performance over the two trials and a second improved 17.8%. A two-way ANOVA (condition × time) with repeated measures on both factors showed no significant main effect differences for condition (F1,6=3.93 P > 0.05), time (F1.83,48=1.12, P > 0.05) or interaction between them (F1.93,48, P > 0.05). Furthermore, neither heart rate nor blood lactate responses in the boxing task differed between trials. These data were affected by the small sample. Power and effect size analysis using η2 procedure and removing the outlier data produced a mean fall in boxing performance of 26.8%. However, some subjects appeared able to resist the deleterious effects of a rapid loss of body mass prior to competition and further research is needed to explain the mechanisms under-pinning this ability.