One thousand sixteen gastric cancer (GC) patients and 1623 population controls, interviewed in a multicentric study in Italy, reported their family history for gastric, esophageal, and colorectal cancer. A significant association was found only with a history of GC in a sibling or in a parent [odds ratio (OR), 2.6 and 1.7, respectively], which persisted after adjusting for potential confounders including nutrient intake. Adjusted GC risk was higher for subjects having an affected mother than an affected father (OR, 2.3 and 1.3) and showed a further increase for subjects reporting both parents (OR, 3.0) or two or more siblings affected with GC (OR, 8.5). The proportion of patients with an affected first-degree relative was higher among females, in the elderly, and in high-risk areas. Among adult siblings of controls and cases, GC prevalence reported at interview was 1 and 2.7%, respectively; a further increase was shown in families with at least one parent affected (1.4 and 5.7%). GC risk associated with a positive family history was greater among residents of low-risk areas where risks were increased about two-fold. Among cases, family history of GC was not related to blood group A or to histological type according to the Lauren classification. These findings are discussed in terms of the contributions of genetic and environmental risk factors and their possible interactions.