The lactate threshold is a widely used and, at times, controversial construct in exercise physiology and pathophysiology. Its non-invasive estimation during incremental exercise depends upon CO2 output increasing as a function of O2 uptake, i.e. 'V-slope', as a result of bicarbonate buffering during the lactic acidosis. However, we hypothesised that the V-slope deflection could also occur as a consequence of metabolic CO2 being diverted proportionally more into the CO2 stores in the early phase of exercise. Eight healthy males performed two incremental exercise tests on a cycle ergometer, with and without controlled prior hyperventilation; the hyperventilation caused end-tidal PCO2 to decline by [sim]10 mmHg, with the clearance of a CO2 volume averaging 2547 ml. This corresponded to an 'effective' CO2 capacitance of some 3·12 ml mmHg-1 kg-1. Gas exchange was determined breath-by-breath, and blood was sampled from the dorsum of the heated hand. Our results demonstrate that the early dynamics of CO2 wash-in to the previously depleted body stores can result in a 'pseudo-threshold', i.e. significantly before the onset of the actual lactic acidosis. Precautions should therefore be taken to avoid hyperventilation prior to non-invasive estimation of the lactate threshold.