Activating the upper body during walking, Nordic walking (NW) may be used as an external cueing to improve spatiotemporal parameters of gait, such as stride length or gait variability, in Parkinson disease (PD). Structured gait variability, revealed by the presence of long-range autocorrelations (LRA), was associated to dynamic stability of gait. Dynamic stability has been defined as the ability to maintain functional locomotion despite the presence of internal or external disturbances, which is a feature of healthy locomotor system. This study aimed to study beneficial effects of NW on dynamic stability of gait in PD. After three sessions of practice, 14 mild to moderate PD patients performed 2 × 12 min overground walking sessions (with and without pole in a randomized order) at a comfortable speed. Gait cadence, gait speed, stride length and temporal organization (i.e. LRA) of stride duration variability were studied on 512 consecutive gait cycles using a uni-dimensional accelerometer placed on the malleola of the most affected side. The presence of LRA was based on scaling properties of the series variability (Hurst exponent) and the shape of the power spectral density (α exponent). In order to assess beneficial effects of NW on PD gait, a paired t- test was used ( p < 0,05). All patients presented LRA in all series of walking pattern. However, Hurst and α exponent were significantly higher during NW ( p ≤ 0.001). While gait speed remained unchanged between two walking sessions ( p = 0.320), gait cadence decreased and stride length increased significantly ( p = 0.009 and 0.003 for gait cadence and stride length, respectively). This study demonstrates the presence of LRA during Nordic walking and that way of walking can improve the dynamic stability of gait in Parkinson disease. Such improvement could be due to the upper body rhythmic movements acting as rhythmical external cue to bypass their defective basal ganglia circuitries. Therefore, Nordic walking may constitute a powerful way to manage gait disorder in PD.