Tuberous Sclerosis Complex (TSC) is a single gene disorder – caused by mutation in the TSC1 or TSC2 gene – that carries a high risk of autism spectrum disorder (ASD). Various neurological complications increase the risk of ASD but the way risk factors operate together is unclear. We aimed to explore risk pathways to ASD by modelling the interplay between genetic mutation (TSC1/TSC2), cortical tuber count, seizure type and severity. The Tuberous Sclerosis 2000 Study is a UK population-based, prospective study of the natural history of TSC. We recruited newly diagnosed children (N=125, 49.6% male, median age=2.7 years [range 4 weeks – 18 years]) and collected data on mutation, cortical tuber count (cranial MRI/CT), seizure history, and IQ. ASD and IQ assessments were carried out at 10-year follow up (N=86, 45.0% male, median age=12.5 years [range 7.8 – 26.9 years]). Assessment for ASD included the Autism Diagnostic Interview–Revised (ADI-R) and the Autism Diagnostic Observation Schedule (ADOS-2). Structural equation modelling using full information maximum likelihood estimation was used to explore pathways that mediate between mutation and ASD. Risk of ASD was high: 39.5% met research criteria for ASD and a further 41.9% showed autistic traits. Structural equation modelling resulted in two indirect pathways, with cortical tuber count and occurrence/severity of epileptic spasms in infancy mediating between mutation and ASD (mutation→tubers→spasms→ASD, B=2.08, 95% CI 0.15–8.02; mutation→spasms→ASD, B=2.98, 95% CI 0.04–8.89). Concurrent seizures (B=3.08, 95% CI 0.42–6.18) and IQ (B=-117.10, 95% CI -183.57–-59.16) were also associated with ASD symptoms. There was significantly elevated risk of ASD and subclinical autistic traits. Tuber count and severity of epileptic spasms predicted ASD severity, suggesting that seizures in infancy may push genetically vulnerable individuals over the threshold for an ASD diagnosis. Prevention/control of seizures in infancy may decrease severity of ASD symptoms. However, ASD was occasionally reported in the absence of overt seizures in infancy, so their causal role requires further investigation.