Background. Muscle dysfunction is an early complication of diabetic peripheral neuropathy (DPN). As a convenient and low-cost tool, muscle ultrasound has been used to assess muscle quality and muscle mass. However, the relationship between different muscle ultrasound parameters and DPN is unclear. Objectives. This study was designed to investigate the relationship between ultrasound parameters of different muscles and DPN among patients with type 2 diabetes mellitus, including the rectus femoris (RF), tibialis anterior (TA), and medial head of gastrocnemius (MG). Materials and Methods. The research enrolled 90 patients with type 2 diabetes mellitus (T2DM). All images were attained from both sides. Muscle measurements contained muscle thickness (MT), cross-sectional area (CSA), echo intensity (EI), and corrected EI. The binary logistic regression and multiple linear regression were used to investigate the association between muscle ultrasound parameters and DPN or vibration perception threshold (VPT). Results. EI, corrected EI, MT of MG, and EI of TA were associated with DPN separately after adjusting other clinical variates. Among these muscle parameters, the EI of MG had a better predictive value (OR: 1.114, 95% CI: 1.039, 1.196) of DPN. Combined with CSA of RF, peripheral artery disease (PAD), and sex, the corrected EI of MG was associated with the vibration perception threshold (VPT) (standard
β
=
0.242
,
p
<
0.001
), better than the EI of MG (standard
β
=
0.215
,
p
=
0.002
). Conclusions. MG (MT, EI, and corrected EI) and TA (EI) were associated with DPN, respectively. CSA of RF and corrected EI or EI of MG combined with PAD and sex were associated with VPT significantly, which supported that muscle ultrasound might be a substantial quantitative tool for detecting the exercise benefits for DPN.