Numerous reports exist concerning relationships between muscle fiber types or myosin heavy chain expression and in vivo human muscle performance. Whereas in vitro high power performances in single muscle cells are most associated with type IIB/X fibers expressing primarily MHC IIb/x, these relationships are not always observed for in vivo volitional performances in humans. Indeed, previous data suggest that high power performances in humans may be correlated with other MHC isoforms (i.e., MHC I or IIa), depending on prior training experience. Comparisons of data are somewhat difficult when different measures of power or different assessments of fiber characteristics are used. Additionally, training experiences of subjects are sometimes difficult to ascertain when comparing different studies. To compare relationships between each MHC isoform and muscle power for subjects with differing training histories. Healthy adult men with a variety of training backgrounds served as subjects for this study (n = 37, [X ± SD] age = 23.3 ± 4.0 yrs., hgt. = 178.4 ± 6.3 cm, BW = 84.8 ± 21.4 kg, % fat = 12.5 ± 6.7 %). Subjects were divided into three subject groups; elite lifters (ELITE; n = 5, 10 yrs. trng experience, 4-5 training sessions·wk−1), moderately weight trained (MOD; n = 27, 1-3 yrs. training experience, 2-3 training sessions·wk−1), and sedentary (SED; n = 5, no training ≥3 yrs.). All subjects performed depth vertical jumps (DVJ) from a height of 0.9 m from which peak power (W) was estimated. Muscle biopsies were obtained from the vastus lateralis m. and analyzed via SDS-PAGE to determine relative expression of myosin heavy chain (MHC) isoform expression. Regression analyses were performed for each subject group between each MHC isoform (i.e., I, IIa, IIb/x) and DVJ peak power (p.81), MOD = MHC type IIb/x (r2>.29), and ELITE = MHC type IIa (r2>.44). The ability to optimize fiber recruitment for a high power movement such as a DVJ is not achieved until large volumes of training have occurred. Untrained individuals most readily recruit the low threshold motor units that contain primarily MHC type I. Moderately trained individuals who still express some MHC IIb/x can recruit these motor units when necessary for activities such as a DVJ. Highly trained individuals who express little or no MHC IIb/x depend preferentially on motor units containing MHC type IIa. This trainingdependent continuum of fiber contributions to high power performance indicates that optimizing high power performance depends on intense, long-term training programs (e.g., >1-3 yrs., >2-3 sessions·wk−1). Funded by grants from the NSCA, the Fed Ex Technology Institute, and GNC, Inc.