Background: Body weight has been recognized as a driving factor of osteoarthritis. Few studies had investigated the association between weight status across adulthood and risk of osteoarthritis (OA). This study investigates the association of weight change patterns across adulthood (lasting at least 25 years) with the risk of OA from the National Health and Nutrition Examination Survey (NHANES) 2013–2018. Methods: The study assessed the relationship between weight change across adulthood and OA in 7392 individuals aged > 50 spanning a minimum of 25 years. Multivariate linear regression analyses were utilized to detect the association between weight change patterns and self-reported OA. Restricted cubic splines (RCS) were used to examine the nonlinear relationship between absolute weight change and OA risk. Results: From 10 years ago to survey, the risk of OA was 1.34-fold (95% CI 1.07–1.68) in people changed from obese to non-obese, 1.61-fold (95% CI 1.29–2.00) in people change from non-obese to obese, and 1.82-fold (95% CI 1.49–2.22) in stable obese people compared with people who were at stable normal weight. Similar patterns were also observed at age 25 years to baseline and age 25 years to 10 years before the baseline. The dose–response association of RCS found a U-shaped relationship between absolute weight change and OA risk. Conclusions: The study suggests that weight patterns across adulthood are associated with the risk of OA. These findings stressed important to maintain a normal weight throughout adulthood, especially to prevent ignored weight gain in early adulthood to reduce OA risk later.