Background: In the present study, we investigated the clinical significance of serum soluble IL-2R as a predictor of lymph node metastasis in patients with early gastric cancer. Patients and Methods: Seventy-four patients with early gastric cancer were enrolled in this study. Levels of serum soluble IL-2R were measured by an enzyme-linked immunosorbent assay. Result: Significant differences in serum soluble IL-2R between the control group and cases of T1 were not recognized. On the other hand, levels of serum soluble IL-2R in 74 patients with early gastric cancer (T1) were significantly lower than those of T2, T3, and T4 (p < 0.05). There were no significant differences in serum soluble IL-2R between cases of mucosal and submucosal invasion (379 ± 42 vs. 382 ± 35 U/ml). Six of 35 patients with submucosal invasion (17.1%) had lymph node metastasis, but none of the 39 patients with mucosal invasion. In the 6 cases showing lymph node metastasis, the macroscopic types were IIc + Ul(+) in 4, and IIc + IIa and IIc + IIb in 1, respectively. Histopathologically, there were 5 poorly and 1 moderately differentiated adenocarcinomas. In 6 cases with lymph node metastasis, serum soluble IL-2R levels were significantly higher than in those without lymph node metastasis (556.8 ± 73 vs. 329 ± 22 U/ml, respectively, mean ± SEM, p < 0.05). Five of these 6 cases demonstrated statistically significantly increased levels of serum soluble IL-2R (sensitivity 83%, specificity 63%), suggesting serum soluble IL-2R as a predictor of lymph node metastasis in early gastric cancer (p < 0.05). Conclusion: According to these data, in patients of early gastric cancer with increased levels of serum soluble IL-2R, endoscopic mucosal resection or minimal invasive gastrectomy without dissection of regional lymph nodes should be avoided, since there is a high risk of lymph node metastasis.