Barrett's esophagus (BE) is a precursor lesion for esophageal adenocarcinoma/gastric cardia adenocarcinoma (EA/GCA), which are cancers with increasing incidences in the United States (US) and very poor prognoses. In experimental studies, the impact of polyunsaturated fatty acids (PUFAs) on carcinogenesis varies, with o-3 PUFAs (primarily found in fish) and o-6 PUFAs (often found in oils and other foods) displaying anti- and pro-carcinogenic effects, respectively. My hypotheses were that the risk of developing BE/EA/GCA and/or dying from EA/GCA would be inversely associated with non-fried fish intake and other measures of o-3 PUFAs, but positively associated with o-6 PUFAs. In my dissertation, I pooled two case-control studies of BE and two case-control studies of EA/GCA with case follow-up for mortality. The total sample size included 471 BE cases with 490 controls, 1027 EA/GCA cases with 2027 controls, and 884 EA/GCA deaths. Using study-specific food frequency questionnaires, I harmonized and pooled dietary information to estimate PUFA measures including intake of fish (with consideration given to cooking methods), o-3, o-6, and o-6:o-3 ratio. Using logistic, polytomous logistic, and Cox proportional hazards regression models, I estimated odds ratios and hazards ratios, respectively, with 95% confidence intervals. Higher intake of baked/broiled fish was associated with approximately 30% decreased risk for development of BE (particularly the more severe long-segment BE), EA, and GCA. o-6 intake was also associated with an increased risk of EA and GCA; however, so was o-3. Finally higher o-6:o-3 was associated with lower EA mortality, but not GCA mortality. There was no evidence of modification by inflammation-related factors for any of the outcomes assessed. My findings of inverse associations for baked/broiled fish intake with BE/EA/GCA development and positive associations for o-6 with EA/GCA development are consistent with my hypotheses. But the positive association between o-3 intake and EA/GCA development, and the inverse association between o-6:o-3 and EA mortality, are not. If findings are confirmed, increasing intake of baked/broiled fish may be a plausible risk reduction strategy for BE (especially long-segment BE), EA, and GCA, and could reduce the disease burden of these lethal cancers.