According to World Health Organization, unipolar depressive disorders were ranked as the third leading cause of global burden in 2004 and predicted to move into the first place by 2030. Bipolar Disorder affects approximately 5.7 million adult Americans every year (NIHM). Bipolar Disorder (BD) is recognized as a chronic mental illness, which requires consistent monitoring. Due to the unpredictability of this disease, any attempts to manage BD are limited to constant tracking of the patients’ vitals, monitoring and auditing the behavior and the triggers in the environment. This thesis develops an algorithm as a solution for predicting the onset of a bipolar episode with the use of mobile technology. Several factors can lead to increase in stress levels, which is a major cause for the onset of a bipolar episode. Variables such as sleep, mood and heart rate have been reported to have a direct impact on stress levels. The prediction algorithm has been designed and developed for the iOS mobile platform in which various variables responsible for the onset of a bipolar episode are taken into account. Wearable sensor has been used to capture the heart rate information. Mood and sleep data is being obtained via a self-reporting mechanism where the application prompts the user to provide the input. This data is then used to identify a bipolar episode and generate alerts depending upon the prediction algorithm.