Autophagy is the process by which portions of cytoplasm are enclosed by membranous organelles, phagophores, which deliver the sequestered cytoplasm to degradative autophagic vacuoles. Genes and proteins involved in phagophore manufacture have been extensively studied, but little is known about how mature phagophores proceed through the subsequent steps of expansion, closure and fusion. Here we have addressed these issues by combining our unique autophagic cargo sequestration assay (using the cytosolic enzyme LDH as a cargo marker) with quantitative measurements of the lipidation-dependent anchorage and turnover of the phagophore-associated protein LC3. In isolated rat hepatocytes, amino acid-starved to induce maximal autophagic activity, the two unrelated, reversible autophagy inhibitors 3-methyladenine (3MA) and thapsigargin (TG) both blocked cargo sequestration completely. However, whereas 3MA inhibited LC3 lipidation, TG did not, thus apparently acting at a post-lipidation step to prevent phagophore closure. Intriguingly, the resumption of cargo sequestration seen upon release from a reversible TG block was completely suppressed by 3MA, revealing that 3MA not only inhibits LC3 lipidation, but also (like TG) blocks phagophore closure at a post-lipidation step. 3MA did not, however, prevent the resumption of lysosomal LC3 degradation, indicating that phagophores could fuse directly with degradative autophagic vacuoles without carrying cytosolic cargo. This fusion step was clearly blocked by TG. Furthermore, density gradient centrifugation revealed that a fraction of the LC3-marked phagophores retained by TG could be density-shifted by the acidotropic drug propylamine along with the lysosomal marker, cathepsin B, suggesting physical association of some phagophores with lysosomes prior to cargo sequestration.