Background: Postoperative infectious complications (PICs) after gastric cancer resection remain a clinically relevant problem. Early detection of PICs, before critical illness develops, may be of considerable clinical benefit. The aims of this study were to investigate the predictive factors for PICs and to define the clinical parameters for their early detection in patients with gastric cancer resection. Methods: Clinical data of 417 consecutive patients undergoing elective gastrectomy for primary gastric cancer between 2009 and 2012 were retrospectively analyzed. Diagnostic accuracy was determined by the area under the receiver operating characteristic curve (AUC). Univariate and multivariate logistic regression analyses identified clinical factors predicting PICs of grade III or more according to the Clavien-Dindo classification. Results: Forty-four patients developed PICs of grade ≥ III (10.6%, 95% confidence interval [CI] 7.6–13.5%). As a systemic inflammatory marker, C-reactive protein (CRP) on postoperative day (POD) 3 had superior diagnostic accuracy for PICs (AUC 0.802, 95% CI 0.735–0.870) with a calculated cut-off value of 17.7 mg/dl, yielding a sensitivity of 0.66 (95% CI 0.524–0.774) and a specificity of 0.84 (95% CI 0.821–0.850). Multivariate analysis identified CRP on POD 3 of 17.7 mg/dl or greater (odds ratio [OR] 8.094, 95% CI 3.568–19.342) as well as clinical stage ≥ II (OR 4.445, 95% CI 1.478–15.881) and operation time ≥ 250 min (OR 3.638, 95% CI 1.449–10.137) as significant predictive factors for PICs after gastrectomy. Conclusions: Elevated CRP levels on POD 3 will help physicians predict postoperative course and facilitate decision-making regarding the prompt, comprehensive clinical search and therapeutic approach for PICs.