Abstract Background Random biopsies are recommended to identify individuals at risk of gastric adenocarcinoma. Cumulative evidence suggests that narrow-band imaging (NBI) can be used to grade gastric intestinal metaplasia (GIM). We aimed to externally validate a classification of endoscopic grading of gastric intestinal metaplasia (EGGIM). Methods Consecutive patients in two centers were submitted to high resolution white-light gastroscopy followed by NBI to estimate EGGIM – a score (0 – 10) resulting from the sum of endoscopic assessments of GIM, scored as 0, 1, or 2 for no GIM, ≤ 30 %, or > 30 % of the mucosa, respectively, in five areas (lesser and greater curvature of both antrum and corpus, and incisura). If GIM was endoscopically suspected, targeted biopsies were performed; if GIM was not noticeable, random biopsies were performed according to the Sydney system to estimate the operative link on gastric intestinal metaplasia (OLGIM; the gold standard). Results 250 patients (62 % female; median age 55 years) were included. GIM was staged as OLGIM 0, I, II, III, IV in 136 (54 %), 15 (6 %), 52 (21 %), 34 (14 %), and 13 (5 %) patients, respectively. All patients with GIM except three were identifiable with targeted biopsies. For the diagnosis of OLGIM III/IV, the area under the ROC curve was 0.96 (95 % confidence interval [CI] 0.93 – 0.98) and by using the cutoff > 4, sensitivity, specificity, and positive likelihood ratio were 89 %, 95 %, and 16.5, respectively; results were similar (91 %, 95 %, and 18.1) when excluding patients with foveolar hyperplasia. Conclusions For the first time, an endoscopic approach was externally validated to determine the risk of gastric cancer without the need for biopsies. This can be used to simplify and individualize the management of patients with gastric precancerous conditions.