OBJECTIVE The aim of this study was to determine the magnitude of ketone production in ultra marathon runners and what affect if any this has on performance. METHODS Participants in the Cliff Young Australian Six Day Race (n=31) provided a prerace urine sample and, then, random urine samples throughout the duration of the event, ranging from 4-20 samples each. Based on urinalysis results, participants were divided into two groups: those who formed ketones (ketone group), and those who did not form ketones or formed ketones only once during a race at the lowest recordable value (non-ketone group). RESULTS The average ketone level of the 22 athletes in the ketone group (value+/-standard deviation: 5.67+/-5.59 mg/dL) was statistically different from 9 athletes who were in the non-ketone group (0.18+/-0.14 mg/dL) (P<0.05). The average distances run for the two groups were 498.09+/-153.99 and 535.6+/-181.08 km, respectively (P=0.56). When average ketone value was compared, excluding runners who did not complete the race, the ketone group (5.88+/-1.37) remained statistically different from the non-ketone group (0.2+/-0.45) (P<0.05). The average distances for those athletes who completed the race were 583.9+/-116.09 and 557.8+/-85.82 km, respectively (P=0.52). CONCLUSIONS We conclude that although two runner sub-populations were revealed, runners who produce ketones and runners who do not make ketones, the level of ketones produced did not affect overall distance run, which is the performance criterion of the race. The nature of this extreme event has illuminated a physiologic difference among ultra marathon runners, and although this difference does not appear to affect race performance, the long-term health consequences are unknown and additional rigorous research is warranted.