Whole cereal grain foods are rich in phytate, a Ca chelator, and could increase the risk of hip fracture. The objective of the present study was to investigate the association between baseline whole grain intake and incident hip fracture. In the present study, 29 192 women who at baseline in 1986 were aged 55-69 years, free of diabetes, and reported a plausible energy intake of 2508-20 900 kJ/d and reported no fracture since the age of 35 years were followed. Hip fracture (n 746) was self-reported in five questionnaires through 2004. Of 1451 hip fractures identified passively by Medicare linkage through 31 December 2004 (Medicare hip fracture), 507 had also been self-reported. Whole grain intake was inversely related to Medicare hip fracture (P trend = 0.02), but it was unrelated to self-reported hip fracture (P trend = 0.27). The hazard ratio in the highest to lowest quintile of whole grain intake for incident Medicare-only hip fracture (n 944) was 0.66 (95% CI 0.53, 0.82) after adjustment for age, energy intake, education, BMI, waist-to-hip ratio, farm residence, physical activity, oestrogen use, smoking, alcohol use, history of cancer and other dietary variables. Medicare-only cases may have failed to self-report due to severe illness; hazard ratio for total mortality after hip fracture was 2.92 (2.37, 3.59) for Medicare-only cases v. Medicare-confirmed self-reported cases. In conclusion, in this cohort, the inverse association between whole grain intake and hip fracture was explained by ascertainment bias. Whole grain intake may increase the ability to respond to a questionnaire and self-report hip fracture, and could reflect less undocumented frailty.