This study examined the effect of (a) creatine supplementation on exercise metabolism and performance and (b) changes in intramuscular total creatine stores following a 5 day supplementation period and a 28 day wash-out period. Six men performed four exercise trials, each consisting of four 1 min cycling bouts, punctuated by 1 min of rest followed by a fifth bout to fatigue, all at a workload estimated to require 115 or 125% VO2,max. After three familiarization trials, one trial was conducted following a creatine monohydrate supplementation protocol (CREAT); the other after 28 d without creatine supplementation, in which the last 5 d involved placebo ingestion (CON). Intramuscular TCr was elevated (P < 0.05) in CREAT compared with the final familiarization trial (FAM 3) and CON. Concentrations of this metabolite in these latter trials were not different. In addition, a main effect (P < 0.05) for treatment was observed for PCr when the data from CREAT were compared with CON. In contrast, no differences were observed in the total adenine nucleotide pool (ATP+ADP+AMP), inosine 5'-monophosphate, ammonia, lactate or glycogen when comparing CREAT with CON. Despite the differences in TCr and PCr concentrations when comparing CREAT with other trials, no difference was observed in exercise duration in the fifth work bout. These data demonstrate that creatine supplementation results in an increase in TCr but this has no effect on performance during exercise of this nature, where the creatine kinase system is not the principal energy supplier. In addition 28 d without supplementation is a sufficient time to return intramuscular TCr stores to basal levels.