Romantic relationships play a critical role in human happiness and well-being, yet predicting their success or failure has long puzzled scientists. This difficulty may come from not having enough detailed information about daily relationship experiences. With modern technology, like smartphones, people now track many parts of their daily lives, such as diet, exercise, and mood. This project uses similar technology to better understand romantic relationships by observing how they change day by day. It applies techniques like those used to forecast the weather or stock market trends to identify relationship patterns. These patterns help scientists learn about the different ways that relationships ebb and flow in daily life and predict which relationships are likely to last and which may not. This project seeks to transform the basic scientific understanding of romantic relationship quality. It uses innovative intensive longitudinal methods with large, heterogeneous samples of individuals and couples in romantic relationship to (1) examine dynamic patterns of change in relationship quality over time and (2) identify whether some relationship patterns are more adaptive than others. For instance, patterns involving quick recovery from negative experiences might be more adaptive than those involving slow recovery. Tracking day-to-day relationship changes over long periods allows for discovery of common patterns that emerge before a relationship ends, improving predictions of re