A balance testing methodology was developed, based on a posture control model which defines relative stability by the degree to which a transient postural perturbation would cause the centre-of-pressure on the feet to approach the limits of the base-of-support. To minimize anticipatory adaptations and to ensure subject safety, the balance test used a small-amplitude continuous random or pseudorandom perturbation. The data were used to identify an input-output model, which was then used to predict large-amplitude transient response. The test perturbation was an anterior-posterior acceleration of a platform on which the subject stood. Pilot experiments were performed to determine appropriate perturbation parameters. Testing of sixty-four normal subjects demonstrated highly significant ageing-related decreases in predicted stability. No significant sex-related differences were found. Predicted stability increased when the subjects were blindfolded. In using the balance test to identify balance-impaired individuals, the blindfolded results and the eyes-open/blindfolded ratio provided higher success rates than the eyes-open results. Depending on the modelling method used, the balance test was able to identify up to three of five vestibular patients and five of five elderly fallers, at a false positive rate of 25% in the normal subjects. Correlations between the balance test results and measures of spontaneous postural sway were weak, except in the normal young adult subjects. Comparison with transient tests showed the balance test to yield reasonably accurate predictions of small-amplitude transient response, but to overestimate the large-amplitude response; however, the transient test results may have been confounded by adaptive effects. It was concluded that the balance test provides a sensitive functional measure of the changes in postural control that are known to occur in ageing. Although it shows promise as a tool for identifying balance impairments, larger numbers of balance-impaired subjects must be tested.