in nonambulatory than ambulatory children with CP. Relying on BMI alone may lead to a misclassification of children with CP into underweight, normal/healthy weight and overweight/ obesity categories. Mathematical models were developed to estimate FMI and FFMI in children with CP from BMI and other easily obtainable measures. Background and Objective(s): Adults with cerebral palsy (CP) experience restrictions in participation, among them employment. Little is known about the course of employment in the long term. This prospective cohort study in adults with CP aimed to describe the course of employment over 14 years, and to identify subgroups at risk for unemployment. Study Design: Prospective cohort study. Study Participants & Setting: Sixty-five adults with CP (51% male; baseline age 21-31y; 25% intellectual impairment; 72% GMFCS level I-III) from a representative Dutch cohort (n=88) of adults formerly known in pediatric rehabilitation care participated. Respondents completed assessments in 1996, 2000 and in 2010. Materials/Methods: Data were collected using self-report questionnaires , by post (1996 and 2010) or interview-format (2000). Primary outcomes were employment status (including sheltered work) and-for those employed-work hours/week. Generalized Estimating Equations (GEE) analyses were performed to study course over time and associated factors. Finally, comparisons with the age-matched Dutch general population were made using one-sample binomial or t-tests. Proportions employed (%), mean (SD) hours/week, and odds ratio (OR) or regression coefficient (B) [95%CI] are reported. Results: Overall, the employment rate in adults with CP was stable over time (38 to 45%, OR 1.10 [0.88, 1.36], p=0.41), but significantly lower than in the general population (75 to 86%, p<0.001). Employment rates were specifically low in adults with intellectual impairment, bilateral CP, and in adults with GMFCS levels IV-V. Work hours/week declined (35.0 (7.9) to 31.2 (10.3), B À1.78 [À3.42, À0.15], p=0.03), specifically among females (32.3 (6.4) to 23.4 (7.4), B À4.78 [À7.21, À2.35], p<0.001). Similar to the Dutch general population, females often worked part-time. Conclusions/Significance: Employment in adults with CP was consistently low compared with the general population, but remained stable up to age 45. However, work hours/week decreased, specifically in females with CP. Adults with more severe intellectual and physical limitations are at risk for unemployment. Employers may better meet the needs of