Objectives <p>Early theories of autism aligned closely with contemporaneous theories of developmental cognitive neuroscience at the time. Since then, autism has developed into its own research field, and links with developmental cognitive neuroscience are fewer. This divergence means our understanding of autism may be built on older neuroscientific theories, and the implications of newer theories in developmental cognitive neuroscience for autism and neurodiversity may not be considered.</p> Methods <p>Without assuming any neuroscientific background of the reader, we describe older theories of developmental cognitive neuroscience, namely nativism and empiricism, and consider their relationship to autism. We then contrast these with the more recent theories of neuroconstructivism and interactive specialisation.</p> Results <p>We argue that applying neuroconstructivism and interactive specialisation to autism and neurodiversity suggests we can view neurodivergent brains as optimised differently, rather than disordered. Deficits are not necessarily required, strengths may be expected, and greater scope for variation in neurotypes arises through the role of environment on development.</p> Conclusions <p>We conclude that neuroconstructivism and interactive specialisation are compatible or even supportive of neurodiversity and create a new neuroscientific dimension with which to explore neurodiversity. However, the potential for the environment to influence autistic development challenges some aspects of neurodiversity and highlights the need for collaborative research and clinical practice with autistic people.</p>