Background: Autistic adults are at high risk of co-occurring mental health disorders, including depression, anxiety, nonsuicidal self-injury, and suicidality. Existing services are not adequately meeting autistic adults' mental health needs. To better inform evidence-based care, there is a need to synthesize and integrate autistic people's perspectives about the mental health-related services they receive. Methods: Using the PRISMA method, 40 studies of first-person experiences for autistic adults about mental health-related services met inclusion for this systematic review. Electronic database searches in PsycINFO and PubMed were conducted up to July 2023. Studies needed to be published, qualitative or mixed-methods, and had to include self-report of autistic adults who had experiences with mental health-related services. Original articles were reviewed, coded, and analyzed in line with a modified critical constructivist grounded theory approach. Results: Twelve studies inquired about participants' experiences and needs with general services, and 28 studies explored participants' experiences with specific services such as specific mental health interventions, support groups, or social-focused interventions. Our analyses of the 40 studies revealed five categories with regard to autistic clients' perspectives on their mental health services: (1) to be more tailored to their individual needs, (2) to provide more accommodations in both format and content, (3) to include more practical support, in addition to emotional support, (4) for providers to have both knowledge of the heterogeneity of autism and humility in their work, and (5) to encourage and embrace a sense of autistic identity and community. Conclusion: Our results offer practical changes that clinicians can make to improve their services and create a better experience for autistic adults. Notably, our analyses highlight that too often the services autistic people receive focus on changing them, rather than on changing or coping with the environment. Diversity in participants was a limitation across our studies (participants were majority White, in a Western-based country, and did not have a co-occurring intellectual disability), and future research should ensure greater diversity in first-person perspectives.