Hippocampal slices often have more synapses than perfusion-fixed hippocampus, but the cause of this synaptogenesis is unclear. Ultrastructural evidence for synaptogenic triggers during slice preparation was investigated in 21-day-old rats. Slices chopped under warm or chilled conditions and fixed after 0, 5, 25, 60, or 180 minutes of incubation in an interface chamber were compared with hippocampi fixed by perfusion or by immersion of the whole hippocampus. There was no significant synaptogenesis in these slices compared with perfusion-fixed hippocampus, but there were other structural changes during slice preparation and recovery in vitro. Whole hippocampus and slices prepared under warm conditions exhibited an increase in axonal coated vesicles, suggesting widespread neurotransmitter release. Glycogen granules were depleted from astrocytes and neurons in 0-min slices, began to reappear by 1 hour, and had fully recovered by 3 hours. Dendritic microtubules were initially disassembled in slices, but reassembled into normal axial arrays after 5 minutes. Microtubules were short at 5 minutes (12.3 ± 1.1 μm) but had recovered normal lengths by 3 hours (84.6 ± 20.0 μm) compared with perfusion-fixed hippocampus (91 ± 22 μm). Microtubules appeared transiently in 15 ± 3% and 9 ± 4% of dendritic spines 5 and 25 minutes after incubation, respectively. Spine microtubules were absent from perfusion-fixed hippocampus and 3-hour slices. Ice-cold dissection and vibratomy in media that blocked activity initially produced less glycogen loss, coated vesicles, and microtubule disassembly. Submersing these slices in normal oxygenated media at 34°C led to glycogen depletion, as well as increased coated vesicles and microtubule disassembly within 1 minute. J. Comp. Neurol. 465:90–103, 2003. © 2003 Wiley-Liss, Inc.