Single cell RNA-seq data places us in an unprecedented position where we are able to examine patterns of variation and importantly co-variation of genes across cells along continuous differentiation trajectories. We recently presented Differential Correlation Across Ranked Samples (DCARS), a statistical method to identify differentially correlated gene pairs across a set of ranked samples, representing either discrete or continuous patterns of group identity. Here, we describe a new approach, scDCARS, a framework for which changes in correlation are examined across a differentiation trajectory. We demonstrate scDCARS with liver developmental data and find key cascading changes in coordination of gene subnetworks including those associated with cell cycle and lipoprotein metabolism. Furthermore, we present scDCARS as part of the DCARS package as well as an interactive Shiny application readily available for scientists’ interrogation with new data. This work provides a unique lens in which higher order interactions among genes can be unpicked and understand the landscape of cell type fate choice.