This report documents progress and results of the project Novel Computational Methods for Predicting Transitions in Spatiotemporal Neurodynamics between Attention and Mind-wandering. We aim to build an exploratory and predictive model of the brain that is sensitive to the transitions between sustained attention and mind-wandering behaviors. Such a predictive model potentially has applications in tracking attention during critical tasks as well as being of medical and diagnostic relevance. Towards this goal, we developed novel methods for characterizing and predicting the spatio-temporal dynamics of the brain at two complementary levels with differing types of spatial and temporal resolution: Level 1. Electroencephalography (EEG) microstates, which are short quasi-stable topographies of brain electrical activity as measured at the scalp, on the order of 80-120 milliseconds. Level 2. Functional Magnetic Resonance Imaging (fMRI) functional connectivity maps, which reveal networks of blood-oxygen-level-dependent (BOLD) activation in distributed brain areas at a slower time scale, on the order of seconds, with high spatial resolution. This research examines and relates the regularities in patterns and sequences of EEG microstates and functional connectivity maps, with the intention of predicting transitions between states in humans of attention and mind-wandering.