In human locomotion the ability to generate and sustain mechanical power output is dependent on the organised variability in contractile and metabolic properties of the muscle fibres that comprise the active muscles. In studies of human exercise we have used a microdissection technique to obtain fragments of single muscle fibres from needle biopsies before and after exercise. Each fibre fragment is divided into two parts. One part is used to characterize the fibre type in respect of the heavy chain myosin isoform expressed. The other part of the fragment is analysed for high energy phosphate concentration (Sant’Ana Pereira et al. 1996). Fibres are classified on the basis of expressing either type I, type IIA, or type IIX myosin heavy chain isoforms. It should be noted however that in the type II population many fibres co-express both IIA and the IIX isoforms and we therefore characterize these fibres on the basis of the degree of co-expression. Moreover while there are significant numbers of fibres expressing only the IIA isoform very few fibres are seen in normal healthy subjects which express only IIX. We were able to show that immediately following 25 s of maximal dynamic exercise, during which power output declined by ~50%, phosphocreatine (PCr) was reduced to zero, or near zero levels in all fibres. ATP was also reduced to 53-34% of resting levels in the type II fibre subgroups, and to ~75% in type I fibres, with concomitant increases in IMP (Sant’Ana Pereira et al. 1996). Subsequently we examined the time course of this dramatic depletion in high energy phosphate using shorter duration cycling exercise (~20 contractions in 10 s; Karatzaferi 2001). In these experiments maximum power ouput decreased by ~23%. Fibre fragments were classified as either type I, IIA, IIAx or IIXa (the latter two classifications of co-expressing fibres having respectively a predominance of type IIA or IIX isoform). Immediately post-exercise PCr content in the four fibre populations decreased to 54, 47, 38, and 41% of resting values. ATP showed no change in type I fibres but decreased to 75, 33, and 30% of resting values in type IIA, IIAx and IIXa fibre groups. There was no detectable IMP in the type I fibres but significant IMP production in type II fibre populations despite the presence of PCr. The results suggest that maximal all-out exercise presented a sequential metabolic challenge to first the type IIX-expressing fibres, then IIA fibres and finally the type I fibres. It is, of course entirely reasonable that during maximal activation those fibre populations with the fastest cross-bridge cycling rates, as determined by myosin heavy chain isoform expressed, will deplete high energy phosphates at the greatest rate resulting in selective fatigue of that population. Thus although the whole muscle mechanical ouput may decrease by only 25% in 20 contractions this may obscure the fact that some fibre populations may be generating very little mechanical ouput while others will be relatively unaffected. The progressive reduction of power during maximal sprint efforts may be interpreted as the cumulative effect of metabolic depletion in successive fibre type populations from IIX to IIXa to IIAx to IIA to I. One important application of the micro-dissection technique is that PCr content may be used as a very sensitive metabolic marker for fibre type recruitment during very short duration concentric, isometric and eccentric exercise (Beltman et al. 2001). There are of course considerable difficulties in quantifying the contribution of different fibre type populations to mechanical output during whole body exercise not least because of the velocity dependence of both power and efficiency. Nevertheless the issue is of considerable interest having as it does an impact on choice of movement cadence and strategies for improving muscle function including during, e.g. functional electrical stimulation and rehabilitation therapy. In a series of experiments since 1981 we have used cycling as an experimental model and believe that there is good evidence to support the view that under normal conditions human type I fibres will be operating around their optimum for maximum power at a pedalling rate of about 60 rev/min, with type IIA fibres, and those expressing increasing proportions of IIX myosin heavy chain isoform having optima at increasing pedalling speeds, and with related optima for efficiency. It must be recognized however that the contractile and metabolic muscle fibre properties can be transformed both chronically (e.g. by electrical stimulation or training), or acutely by exercise itself, consequent upon fatigue or by changes in muscle temperature (see Sargeant, 1999; Ferguson et al. 2002).