The authors examined the relation of recalled weight variability and history of intentional and unintentional weight loss with disease history in 41,837 older women. Lifetime history of disease, current medication use, health-related behaviors, body weight, and intentional and unintentional weight loss episodes of at least 20 lbs (9.1 kg) were assessed by means of two surveys, completed 6 years apart. Weight variability, as measured by the root mean square error around the linear regression line of weight on age at ages 18, 30, and 40 years, was positively related to disease history. Women who reported losing > or = 20 lbs (> or = 9.1 kg) unintentionally between the ages 18 and 39 years were more likely to report a history of disease than were women who had never lost > or = 20 lbs (> or = 9.1 kg) during this age period. Intentional weight loss episodes of > or = 20 lbs (> or = 9.1 kg) between ages 18 and 39 years were also associated with higher cumulative disease prevalence. These results suggest that both unintentional and, to a lesser degree, intentional weight loss may contribute to the observed positive relation between weight loss or variability and disease. Prospective studies are needed to confirm whether weight variability is a risk factor for disease only when unintentional, or whether intentional weight loss also increases risk.