During dynamic exercise, the response of the cardiorespiratory system is structured to maintain homeostasis at the cellular level. The rate at which homeostasis is established is largely dependent on the system's structural and physiological integrity. Evidence suggests that any impairment in the functioning of the system might be reflected in a determination of oxygen uptake (VO[2]) kinetics. The kinetics of VO[2] have been quantified in response to step, impulse, ramp and sinusoidal changes in work rate (WR). An alternative approach uses a technique in which the WR is perturbed according to a pseudorandom binary signal. Pseudorandom binary sequence (PRBS) WR forcings have the advantage of being able to provide a determination of vo2 kinetics from a single test session of ~30 min duration. The assessment of VO[2]kinetics using PRBS WR forcings demands that the controlling process behaves in a linear manner. To minimise the contribution of non-linear influences, changes in work intensity must be constrained to the sub-lactate threshold domain. When examining clinical, untrained or young subjects, the necessary reduction in the upper work limit of a PRBS forcing can effect a fall in the distribution of power across the bandwidth of the sequence. If the distribution of power should fall below a critical level, then it can become difficult to elicit discernible responses from the forcing. To resolve this problem, this thesis investigated the potential for developing a multifrequent WR forcing altered to enhance identification of the underlying VO[2] response. The multifrequent WR forcing developed for use in this thesis took the form of a binary sequence. Binary transitions were determined according to a specially constructed multifrequent signal. Signal construction involved redistributing the available signal power to specific harmonics in a chosen range of frequencies. To validate estimates of VO[2] kinetics derived from the multifrequent binary sequence (MFBS) WR forcing, comparisons were made with the data obtained from an established PRBS forcing. When comparing physiological data, it is necessary to consider the amount of variability between trials. Therefore, prior to assessing the agreement between data obtained from the MFBS and PRBS methods, this thesis sought to establish the degree of variability in estimates of VO[2] kinetics derived from PRBS exercise tests. The results presented in this thesis show estimates of the mean response time (MRT) of VO[2] derived from the MFBS method to be 46.8 (4.2) s (mean (standard deviation) seconds), compared with 45.2 (5.0) s for the PRBS method. This suggests that the two methods yield comparable determinations of VO[2] kinetics. Supporting evidence is provided by the limits of agreement. These indicate that the maximum difference likely to occur between the MRT obtained from the two methods (-6.5 to +9.6 s) is less than that expected due to variability in the MRT derived from PRBS forcings (-11.6 to +8.0 s). However, the limits also reveal the poor repeatability of VO[2] response data obtained from the PRBS used in the thesis. Consequently, the use of this data to assess the validity of t the MRT derived from MFBS forcings is not recommended. In addition to poor repeatability, the possibility exists that assessments of VO[2] kinetics derived from MFBS WR forcings will also depend on the distribution of power across the harmonic content of the sequence. Therefore, whilst MFBS WR forcings may be suited to the assessment of VO[2] kinetics in subjects with a reduced tolerance to exercise, there remain doubts concerning both the validity of the response data and applicability of the method. Until these issues have been resolved, care would need to be taken when using estimates of VO[2] kinetics derived from MFBS WR forcings to determine the functional state of the cardiorespiratory system.