Abstract Background Ultra-processed food intake has been associated with chronic conditions and mortality. The aim of this study was to assess the relationship between ultra-processed food intake and incident frailty in community-dwelling older adults. Methods Prospective cohort study with 1,822 individuals aged at least 60 years who were recruited during 2008–2010 in Spain. At baseline, food consumption was obtained using a validated computerized face-to-face dietary history. Ultra-processed foods were identified according to the nature and extent of their industrial processing (NOVA classification). In 2012, incident frailty was ascertained based on Fried’s criteria. Statistical analyses were performed with logistic regression and adjusted for the main potential confounders. Results After a mean follow-up of 3.5 years, 132 cases of frailty were identified. The fully adjusted risks of frailty across increasing quartiles of the percentage of total energy intake from ultra-processed foods were the following: 0.04 (0.02–0.05), 0.05 (0.03–0.07), 0.09 (0.07–0.12), and 0.11 (0.08–0.14). Results were similar when food consumption was expressed as gram per day/body weight. Regarding ultra-processed food groups, the highest versus the lowest tertiles of consumption of yogurts and fermented milks, cakes and pastries, as well as nonalcoholic beverages (instant coffee and cocoa, packaged juices, and other nonalcoholic drinks, excluding soft drinks) were also significantly related to incident frailty. Conclusions Consumption of ultra-processed foods is strongly associated with frailty risk in older adults. Substituting unprocessed or minimally processed foods for ultra-processed foods would play an important role in the prevention of age-related frailty. Trial registration NCT02804672.