A growing body of evidence confirms that mobile digital devices have key potentials as assistive/educational tools for people with autism spectrum disorders. The aim of this paper is to outline key aspects of development and evaluation methodologies that build on, and provide systematic evidence on effects of using such apps. We rely on the results of two R+D projects, both using quantitative and qualitative methods to support development and to evaluate developed apps (n=54 and n=22). Analyzing methodological conclusions from these studies we outline some guidelines for an 'ideal' R+D methodology but we also point to important trade-offs between the need for best systematic evidence and the limitations on development time and costs. We see these trade-offs as a key issue to be resolved in this field.