Personal data acquisition using smartphones has become robust and achievable in recent times: improvements in user interfaces have made manual inputting more straightforward and intuitive, while advances in sensing technology has made tracking more accurate and less obtrusive. Moreover, algorithmic advances in data mining and machine learning has led to better a interpretation and determination factors indicative of health conditions and outcomes. However, these indicators are still under-utilized when providing feedback to the user or a health worker. Mobile health systems that can exploit such indicators could potentially deliver precision feedback personalized to the user’s condition and also lead to increases in adherence and improve efficacy. In this book chapter, we will provide an overview of the state of the art in mobile health feedback systems and then discuss MyBehavior, an example of a feedback system that utilizes individual data streams and indicators. MyBehavior is the first personalized system that provides health beneficial recommendations based on physical activity and dietary data acquired using smartphones. The system learns common healthy and unhealthy behaviors from activity and dietary logs, and then prioritizes and suggests actions similar to existing behaviors. Such prioritization is done to promote a sense of familiarity to the suggestions and increase the likelihood of adoption. We also formulate a basis framework for future systems similar to MyBehavior and discuss challenges with regard to transference and adaptation.