The regeneration of the axons of leech Retzius cells was compared following two different methods of axonal severing: (1) a crush of the whole connective that includes the Retzius axon; and (2) photoablation of a small segment of only the Retzius axon. The photoablation was carried out after filling the Retzius cell with Lucifer Yellow (LY). Several tests were carried out to determine whether the photoablation actually severed the axon. These included (1) using the lipophilic membrane probe DiI as an indicator of membrane severance (2) electron microscopic examination of the photoablated axon after filling it with horseradish peroxidase (HRP); and (3) filling the Retzius cell first with HRP, then photoablating, and looking for the disappearance of the HRP in the photoablated region. These and other observations indicated that the photoablated axon was actually severed. Two differences were seen in the regeneration of the Retzius axon after crush versus after photoablation. First, the sprouting following crush was far more disorganized, and included significantly more lateral spread. Second, after photoablation, over 70% of the axons, upon refilling with LY after 3 days or more, showed the newly introduced LY suddenly extending far down the distal segment, indicating that the proximal and distal segments had become reconnected. This was never seen following a crush. The photoablated axons did not pass HRP into the distal segment, suggesting that the reconnection was not by fusion, but perhaps by a gap junction. The results show that axonal regeneration can take a dramatically different form than it does following a standard crush procedure if, instead, the axon is severed in a way that preserves the structural integrity of the surrounding tissue.