Application of Stable Isotopes and Multi Elemental Fingerprints to Verify the Origin of Premium Chinese Hainan Bananas

China is the world’s largest consumer and second largest producer of bananas. This strong domestic demand consistently provides a reliable income for Chinese banana growers. The geographical origin of food is usually associated with product quality and safety, and this is especially noted for Hainan origin-labeled bananas, which are grown offshore on China’s largest tropical island. Hainan banana is recognized as a premium variety within China’s banana market, but there have been recent impacts on branding, profits, and a reduction in income for banana farmers due to the fraudulent in-market substitution of non-Hainan bananas. In this study, stable isotope and elemental chemometric models were used to differentiate bananas grown in Hainan province (HN) from non-Hainan provinces (NHN). The results showed that HN bananas had a specific isotopic and elemental fingerprint compared to NHN bananas. Bananas sampled from HN and NHN regions showed significant differences in δ13C values (HN: −22.2‰ to −27.7‰, NHN: −22.3‰ to −24.3‰), Al content (HN: 0.00 mg/kg to 0.10 mg/kg, NHN: 0.00 mg/kg to 0.02 mg/kg), Na content (HN: 0.00 mg/kg to 0.09 mg/kg, NHN: 0.00 mg/kg to 0.07 mg/kg), and other elements (p < 0.05). Overall, 14 key variables reflecting climate and soil properties were selected from a group of 53 variables to improve a partial least squares discriminant analysis (PLS-DA) chemometric model. The discrimination accuracy of the test set increased from 84.60% to 90.93% after variable reduction. The use of stable isotopes and elements combined with PLS-DA models provided an effective method for distinguishing Chinese HN bananas from NHN bananas and would be useful as a screening or regulatory tool to confirm instances of origin fraud.