To assess the neuroimaging features of Parkinson's Disease and dystonia by transcranial sonography (TCS). 63 Parkinson's Disease patients, 32 dystonia patients and 81 controls underwent TCS in blind manner. The echo of the substantia nigra (SN) was classified into I-V as half quantitative data. The echo of SN≥III was considered to be positively enhanced, the hyperechogenicity were measured and the hyper-substantia nigra/midbrain (S/M) were calculated. The echo of the lentiform nucleus (LN) was classified into I-III as half quantitative data. The echo of LN≥II was considered to be positively increased and were measured. Semi-quantitative analysis: the ratio of the patients with SN≥III was greater in Parkinson's Disease patients (60.32%, 38/63) than in dystonia patients (12.50%, 4/32) and normal controls (11.11%, 8/72, χ2=19.67, 36.22, P<0.01, respectively), the ratio of the patients with LN≥II was greater in dystonia patients (65.62%, 21/32) than in Parkinson's Disease patients (20.63%, 13/63) and in controls (8.33%, 6/72, χ2=18.69, 37.83, P<0.01, respectively). Quantitative analysis:the median and quartile (M/Q) of the SN hyperechogenieity area in Parkinson's Disease patients [0.73 (0.53) cm2] was greater than in dystonia patients [0.56 (0.53) cm2] and in controls [0.44 (0.19) cm2, H=10.05, P=0.007], the S/M in Parkinson's Disease patients was greater [15.7% (11.5%)] than in dystonia patients [(13.8% (14.2%)] and in controls [8.9% (2.9%), H=6.96, P=0.031]. The M/Q of LN hyperechogenieity area in dystonia patients was greater [0.50 (0.33) cm2] than in Parkinson's Disease patients [0.45 (0.22) cm2] and in controls [0.35 (0.17) cm2, H=10.87, P=0.004]. TCS might find the specific hyperechogenicity of SN in Parkinson's Disease patients (60.32%) and hyperechogenicity of LN in dystonia patients (65.63%), which could provide useful informations to distinguish Parkinson's Disease from dystonia.