How do large molecules, including biological therapeutics, reach the cytosol if delivered from outside the cell? Endocytic internalization (e.g., by clathrin-mediated endocytosis or fluid-phase uptake) is the principal route for bringing these molecules into cells, but most internalized cargo remains entrapped until degraded in lysosomes. Establishing potential mechanisms of endosome escape has generated long-standing controversy, largely because of technical limitations that prevent direct detection. Using recently developed 3D imaging modalities, especially lattice light-sheet microscopy (LLSM) and its combination with adaptive optics (AO-LLSM), we have shown that we can observe endosomal damage and repair and link it to events such as endosomal fission and cargo delivery. We will characterize the mechanism(s) of the delivery events and seek ways to regulate the delivery route, with the aim of achieving a full description of how endosomal membrane damage and repair govern endosomal escape. We will develop novel ways to exploit the deep-learning approaches that will be essential for analyzing the huge data sets we now generate (e.g., 3D, diffraction-limited movies of an entire cell in two or three colors at 2 sec intervals for 5 or more minutes). Our long-term goal is to build on mechanistic understanding to work out how to enhance specific release of desired cargo, such as antisense oligonucleotides (ASOs), microRNAs (miRNAs), and small interfering RNAs (siRNAs), or in other circumstances, how to prevent release of deleterious cargo, such as 𝛂-synuclein and lipopolysaccharide (LPS).