The relative effectiveness of tracers in guiding para-aortic lymph node dissection (PAND) in advanced gastric cancer is undefined. In this single-center, prospective study, we aimed to discuss the effectiveness of such tracers.Between January 2015 and January 2016, 90 consecutive patients with stage T4a gastric cancer were evenly assigned to receive 0.2 mL of carbon nanoparticles (a), methylene blue (b), or no tracer (c) injection through no. 12b lymph nodes before PAND.There was no difference in the baseline characteristics between the three groups. Group A vs. B or C had a higher number of dissected lymph nodes (34.1 ± 9.8, 25.5 ± 5.5, and 22.6 ± 3.7; P < 0.001; B vs. C: P =0.321) and no. 16a2/b1 para-aortic lymph nodes (PANs; 11.8 ± 4.8, 7.0 ± 1.2, and 5.5 ± 1.2; P < 0.001; B vs. C: P =0.178) and similar rates of lymph node metastasis (20.9 ± 17.5%, 19.1 ± 15.1%, and 23.6 ± 19.7%; P = 0.511), positive dissected PAN (23.3% [7/30], 16.7% [5/30], and 16.7% [5/30]), surgery duration (252.9 + 35.4, 244.4 ± 29.0, and 250.3 + 29.9 min; P = 0.421), and blood loss (266.7 ± 115.5, 270.0 ± 82.6, and 260.0 ± 116.3 mL, P = 0.933). There was no common bile duct damage by tracer injection, and one case of duodenal stump fistula, one abdominal infection, and two anastomotic leakages in Groups A-C, respectively, were treated successfully.In advanced gastric cancer treatment, carbon nanoparticle injection into no. 12b nodes appears to better trace no. 16a2/b1 PAN.