We first explored whether the ability of subjects to detect the direction of slow ramp imposed movements may be improved by the application of mechanical noise to muscle tendons. Movements were plantar/dorsal flexion of the ankle at 0.04°/s, and the amplitude was just sub-threshold for each subject. A white noise signal (random vibration), low-pass filtered to 100 Hz and distributed uniformly in amplitude, was applied to both the extensor and the flexor ankle muscle tendons with four different mean amplitudes (20, 30, 100, 280 μm). The population of subjects was observed to exhibit clear stochastic-type behaviour: their ability to determine the direction of sub-threshold movements significantly increased when the two lower levels of noise were added and subsequently decreased when the noise magnitude was enhanced. Second, using microneurography, we explored the response of 9 primary muscle spindle afferents and 8 cutaneous afferents to the same imposed movements with and without noise application. While these conditions of ankle mobilisation were too small to induce a response in most of the recorded afferents, two muscle afferents exhibited responses that were characteristic of aperiodic stochastic resonance behaviour: the unit movement response was either triggered or improved by the application of an optimal level of noise. All cutaneous afferents were unresponsive to the imposed movements with or without noise application. We conclude that ankle movement sense can be significantly improved by adding an optimal level of mechanical noise to ankle muscle tendons and discuss the optimisation of the response of movement-encoding receptors that may account for this improvement. The application of a mechanical noise on ankle muscle tendons may constitute a means of improving postural stability in subjects with sensory deficits.