fMRI signals permit inferences about behaviorally-induced brain network function (Diwadkar & Eickhoff, 2021) and computational approaches to fMRI data can augment psychology’s goal of understanding human brain - behavior relationships (Mišić & Sporns, 2016). Here, we examined how associative learning (Büchel et al., 1999) induces coordination dynamics (CD, Bressler & Kelso, 2016), i.e. fluctuations in cooperative relations, between brain regions. fMRI data were contemporaneously acquired while participants (n=78) performed a negatively accelerated learning task (Martin et al., 2024). From fMRI time series (90 anatomically defined regions, Tzourio-Mazoyer et al., 2002), we estimated undirected functional connectivity (uFC, Pearson’s r) in 37 successive partially overlapping time windows [C(_90,2) = 4005 unique cooperative interactions between region pairs (i,j) per window]. Next, for each uFC_(i,j) in each participant’s data, a time series was formed from the 37 successive r values; each represents the CD for that region pair (henceforth, CD_(i,j)). Similarities between 312,390 CD_(i,j) (4005 x 78) were estimated using dynamic time warping (Berndt & Clifford, 1994), before partitioning the set using k-medoids clustering (Kaufman & Rousseeuw, 1994). From our results, clusters representing high CD across the task included brain regions in the anterior and posterior heteromodal cortex (including dorsolateral prefrontal and parietal cortex), the medial temporal lobe and the thalamus. We further show that in a clinical model of dys-connection and altered CD (schizophrenia)(Bressler, 2003), CD were substantially altered. Our results reaffirm the role of coordination dynamics as a central brain mechanism for implementing dynamic behavior.