The signal processing technology of hearing aids has advanced to a state where the standardized test methods, as given in IEC-118 and ANSI-322 are no longer in close relationship to the actual behavior of real life amplifying devices. Most hearing aids in the beginning new millennium do not have a volume control. A hearing aid has to be tested to normative procedures as it is, not after it is set to a special state, or operating mode for testing. Testing is necessary to compare electro-acoustic data to psycho-acoustically collected data. Electro-acoustic test data are needed for the actually working amplifier settings, and such a test condition is not in the current teaching of either the ANSI-322 or the IEC-118 standards. To complicate this, the test signal itself is no longer adequately represented by a pure tone or a beat frequency signal or a steady state noise signal. The actual parameters in a hearing aid as programmed for an individual hearing loss make the hearing aid act as a very nonlinear amplifier. Nonlinear amplification together with highly variable time constants in different signal processing bands may no longer be tested deterministically. Stochastic procedures with specified tolerance bands might give faster results. Nearly all hearing aids come with an option for two microphones to change the directivity pattern of the aids. Test boxes are not able to verify the directivity data. Five new test areas are proposed to expand IEC-118 and/or ANSI-322: a) time delay between input and output, b) discrete limitations in time and in amplitude resolution, c) different nonlinear amplification for speech and noise signals, d) changing microphone directivity to changes in incoming sound patterns, and e) compensation for any distorted perception of the human cochlea the hearing aid is fitted to.