Objective The study attempted to establish a less invasive model for dissecting and monitoring the sympathetic nerve in rats via a dorsal approach for subarachnoid block-related studies. Methods A traditional abdominal approach model and a new dorsal approach model were established in SD rats, and the modeling time of the two models was observed. The stability of the new model was evaluated by measuring blood pressure(BP), heart rate (HR), percentage change in lumbar sympathetic nerve activity(LSNA change%), norepinephrine (NE), and nitric oxide( NO) content after subarachnoid injection of bupivacaine. Results ( 1) Building a new model: the time required to create new models for the dorsal approach (DA)group was shorter than that for the traditional abdominal approach (VA)group, as shown by the result (P<0. 0001). (2)Evaluation of the new model: compared with the NS group, the MAP and SBP were lower at T2(5 min after injection of bupivacaine into the subarachnoid space) and T3 (10 min)(P<0. 05); the LSNA change% was significantly different (P<0. 05); the concentration of NE was lower at T3(P<0. 05). Conclusions The study presents a novel lumbar sympathetic anatomy model using the dorsal approach for subarachnoid-related investigations, which was effectively employed to examine the impact of subarachnoid block anesthesia on lumbar sympathetic nerve activity.