Although evidence suggests significant cross-sectional relationships between attention-deficit/hyperactivity disorder (ADHD) and several physical health conditions, less is known about their longitudinal associations. We investigated the cumulative effect of childhood physical health conditions on ADHD symptoms at age 17 years, controlling for environmental factors, ADHD medication status, and ADHD symptoms at age 3 years. Using Millennium Cohort Study data (weighted n = 8,059), we assessed whether 4 physical health clusters (sensory, neurological, atopic, and cardio-metabolic) were associated with scores on the ADHD subscale from the Strengths and Difficulties Questionnaire at age 17 years. Environmental factors were grouped into 5 cumulative risk indices: prenatal, perinatal, postnatal environment, postnatal maternal well-being, and sociodemographic factors. Regression analyses determined whether each physical health cluster was associated with ADHD score while controlling for environmental factors, ADHD medication, and earlier symptoms. Sensory, neurological, and cardio-metabolic clusters were all significantly associated with ADHD symptoms (β range = 0.04-0.09, p < .001). The overall model explained 2% of the variance. This rose to 21% (ΔR = 0.06) after adjusting for confounders. The sensory (β = 0.06) and neurological (β = 0.06) clusters remained significant (R = 0.21, ΔR = 0.06), but the cardio-metabolic cluster was no longer a significant predictor. Sensory or neurological conditions in childhood were associated with higher ADHD symptoms aged 17 after adjustment of confounders. This was not the case for atopic or cardio-metabolic conditions. These findings have implications for the care of children with sensory/neurological conditions and future research examining ADHD etiopathophysiology.< Réduire