Determination of the strongest possible relationship between isokinetic quadriceps and functional performance measurements in healthy females would allow sports medicine practitioners to establish normative values when examining muscular performance in injured females. Previous attempts to correlate both measurements have, however, produced inconsistent results. The purpose of this study was to examine the effects of allometric scaling, isokinetic testing velocities, reciprocal and non-reciprocal isokinetic testing on the relationship between countermovement jump (CMJ) and isokinetic quadriceps torque and power in recreational females athletes. Seventeen females (age 21.0 +/- 2.0 years, body mass index 19.5 +/- 1.0 kg x m(-2)) performed isokinetic quadriceps and CMJ tests. Isokinetic peak torque and average power were obtained reciprocally and non-reciprocally at 1.05 and 3.14 rad x s(-1), and were corrected for body mass by allometric modelling. Pearson product-moment correlation (r) was used to assess the relationship between the isokinetic parameters and the CMJ measurements. Coefficients of determination (r(2)) were calculated to determine the magnitude of common variance. The r-values for all non-allometrically modelled non-reciprocal parameters were greater (r = 0.58-0.63) than isokinetic parameters obtained reciprocally (r = 0.28-0.47). Using allometric scaling, non-reciprocal isokinetic data accounted for an additional 2-9% of the CMJ height variance, and statistically significant correlations were obtained at both 1.05 and 3.14 rad x s(-1). Allometrically scaled, non-reciprocal isokinetic peak torque and average power at 1.05 rad x s(-1) had the highest correlation with CMJ (r(2) = 0.49). At both 1.05 and 3.14 rad x s(-1), non-reciprocal quadriceps parameters correlated more closely with CMJ measurements than do reciprocal contractions. Normalization for body size by allometrically scaling may further improve correlations with CMJ performance.