Autism researchers can be ableist, including by talking about autistic people in sub-human terms (dehumanisation), treating autistic people like objects (objectification), and making othering statements which set autistic people apart from non-autistic people, and below in status (stigmatisation). This mixed-method study aimed to investigate how autism researchers construct autistic people and autism research, and to investigate whether including autistic people more in research relates to lower ableism in narratives about autistic people. We used a survey with autism researchers (N =195) asking five open-ended questions about autism and autism research, as well as demographics, career length, contact with autistic people (familial and non-familial) and degree to which researchers involve autistic people in their research. We used content analysis to categorize narratives used by autism researchers and cues for ableism (dehumanization, objectification, and stigmatisation). We then used binary-logistic regression to identify whether narrative or higher inclusion of autistic people predicted fewer ableist cues, controlling for career length and connections to autistic people. Using medicalised narratives of autism predicted higher odds of ableist cues compared to employing social model or neutral embodiment narratives. Greater inclusion of autistic people in research predicted significantly lower odds of ableist cues, while controlling for other contact with autistic people and career length. Next, we used reflexive thematic analysis to analyse researcher’s perceptions of autistic people and autism research. Narratives reflected core ideological disagreements of the field, such as whether researchers consider autism to be an intrinsic barrier to a good life, and whether researchers prioritise research which tackles “autism” versus barriers to societal inclusion for autistic people. Instrumentality (a form of objectification) was key to whether researchers considered a person to have social value with emphasis revolving around intellectual ability and independence. Lastly, language seemed to act as a tool of normalisation of violence. Researchers relied on an amorphous idea of “autism” when talking about prevention or eradication, potentially because it sounds more palatable than talking about preventing “autistic people”, despite autism only existing within the context of autistic people.