Aim The prevention of pneumonia is an urgent issue among Japanese older adults. However, little has been reported on the relationship between a Functional Independence Measure (FIM) and the Geriatric Nutrition Risk Index (GNRI) for the prevention of pneumonia in patients in long‐term care facilities in Japan. We aimed to clarify the relevance of FIM and GNRI for inpatients with and without pneumonia. Methods We identified 233 patients who were hospitalized in our long‐term nursing hospital from April 2012 to September 2013. We compared differences in FIM among GNRI classes for four groups: (i) pneumonia/high GNRI; (ii) pneumonia/low GNRI; (iii) no pneumonia/high GNRI; and (iv) no pneumonia/low GNRI. To assess the pneumonia predictors, we used a logistic regression for long‐term nursing patients. Receiver operating characteristic analysis showed cut‐off values and the area under the curve. Results A total of 88 (37.8%) of 233 inpatients had pneumonia. FIM of the pneumonia/low GNRI groups was significantly lower than that of the no pneumonia/high and low GNRI groups. Logistic regression showed that FIM (P < 0.001; OR −1.035, 95% CI −1.019–1.051) and GNRI (P = 0.017; OR −1.038, 95% CI −1.007–1.070) were predictors of pneumonia. The cut‐off values for FIM and GNRI were 26.6 (P < 0.001, the area under the curve 0.70) and 80.5 (P < 0.001, the area under the curve 0.65), respectively. Conclusion Low activity and malnutrition might lead to the development of pneumonia. FIM and GNRI are useful predictor tools that could help to prevent pneumonia in Japanese patients in long‐term care facilities. Geriatr Gerontol Int 2016; ••: ••–••.