Purpose The purpose of this study is to establish a new method to reduce the radiation dose during puncture and cannulation in percutaneous endoscopic lumbar discectomy (PELD). Methods Sixty patients with lumbar disk herniation undergoing PELD were prospectively enrolled and randomly divided into an ultrasound (US) guidance group and an X-ray guidance group. The puncture, cannulation, and total operation times; number of fluoroscopy shots; and radiation dose were recorded in both groups. The factors influencing the operation were analyzed. The clinical effect of PELD was evaluated using the straight leg elevation test, visual analog scale (VAS) and Oswestry disability index (ODI). The researchers who collected and analyzed the data were blinded to the group assignments. Results The puncture, cannulation and operation times in the US group were comparable to those in the X-ray group. The patients in the US group received 2.13 +/- 0.35 fluoroscopy shots and a radiation dose of 5.34 +/- 0.63 (mSV), which were significantly lower than the values in the X-ray group (7.57 shots +/- 2.99 shots and 18.25 mSV +/- 10.52 mSV) (P < 0.001). In the US group, the puncture time was significantly longer at the L5-S1 level, in patients with a BMI greater than 28 kg/m(2) and in patients with a high iliac crest. The US and X-ray groups had comparable VAS and ODI scores 1 h and 3 months after PELD, and the VAS scores were significantly lower after PELD (all P < 0.001). No complications were observed in either group. Conclusions US guidance is a new method that reduces the radiation dose required during puncture and cannulation in PELD. [GRAPHICS] .