Testing automated oscillometric sphygmomanometers using simulators is important to ensure good operation. Simulators, however, are unable to model certain aspects of the human hemodynamics. To identify these aspects, the technique of wavelet analysis is used to decompose a set of 5 real and 5 simulated signals. Both continuous (CWT) and discrete (DWT) wavelet transforms are tested. The detection efficiency of various mother wavelets is also tested. Using a CWT with Morlet mother wavelet, the simulated signals whose nature is repetitive and nested are easily identified. A difficulty when trying to use DWT was encountered, in that most of these characteristic human variations occur at scale coefficients other than powers of 2. The obvious differences between simulated and real waveforms show that there are still aspects of the human system which are not properly simulated by machines. The wavelet analysis technique has potential to help isolate and model these aspects.