OBJECTIVE: Throughout the world, the proportion of the male population aged 65 years and older is increasing. Yet, we have limited information regarding diet quality and predictors of diet quality in this segment of the population. The objectives of the current analyses are to describe the diet quality of a cohort of men >65 years of age, and identify lifestyle factors associated with poor diet quality. METHODS: We present a cross-sectional analysis of the diet quality of 5928 men, aged 65-100 years, who are participants in the Osteoporotic Fractures in Men (MrOS) cohort study. Dietary intake was determined using a modified Block 98 food-frequency questionnaire. Diet quality was calculated using the previously validated Diet Quality Index-Revised (DQI-R). Univariate and multivariate modelling was used to estimate the variance in diet quality predicted by a number of sociodemographic factors, including age, race/ethnicity, body mass index (BMI), marital status, education, smoking status, physical activity, self-perceived health and nutritional supplement use. RESULTS: Overall, we found that in this geographically diverse group of older men, diet quality was low, with a mean modified DQI-R for the entire study population of 62.5 (standard deviation 13.1) out of an ideal of 100. Further, younger age, very low total calorie intake (< or = 1187 kcal day- 1), higher BMI, residence in a North or Southeast community, being of African-American or Hispanic race, being less educated, not using dietary supplements and smoking were each significant independent predictors of a poorer diet. CONCLUSION: These data may prove useful in both understanding the dietary intake of older US men as it relates to published dietary guidelines, and for targeting future dietary intervention programmes.