KeepCalm is a digital mental health application, co-designed with community partners, that incorporates wearable biosensing with support for teams to address challenging behaviors and emotion dysregulation in children on the autism spectrum. We followed a user-centered design framework. Before app development, we conducted design workshops, needs assessment interviews, a systematic review, and created an Expert Advisory Board. Once we had a working prototype, we recruited 73 participants to test and help improve the app across five testing cycles. Participants rated the app across testing cycles as highly acceptable, appropriate, feasible, and with good usability. Qualitative data indicated that KeepCalm helped teachers (a) be aware of students' previously unrealized triggers, especially for nonspeaking students; (b) prevent behavioral episodes; (c) communicate with parents about behaviors/strategies; and (d) equipped parents with knowledge of strategies to use at home. We learned that in order to make the app acceptable and appropriate we needed to make the app enjoyable/easy to use and to focus development on novel features that augment teachers' skills (e.g., behavioral pattern and stress detection). We also learned about the importance of maximizing feasibility, through in-person app training/support especially regarding the wearable devices, and the importance of having aides involved. Our findings have informed plans for wider-scale feasibility testing so that we may examine the determinants of implementation to inform adaptations and refinement, and gather preliminary efficacy data on KeepCalm's impact on reducing challenging behaviors and supporting emotion regulation in students on the autism spectrum.