This study examines the effects of several experimental compounds [melatonin (MEL), cyclosporin A (CsA), glial-derived neurotrophic factor (GDNF), and methylprednisolone (MP)] on polyethylene glycol (PEG)-induced repair in vitro and/or in vivo by plasmalemmal fusion (PEG-fusion) of sciatic axons severed by crushing. As measured by conduction of compound action potentials (CAPs) through the lesion site, a significantly (p<0.025) higher percentage (75%) of crushed rat sciatic axons can be repaired in vitro by PEG-fusion following exposure to MEL compared to PEG-fusion of severed sciatic axons in control Krebs saline that contains calcium (CTL=20%). In contrast, no other experimental compound (GDNF: 45%; MP: 42%; CsA: 24%) produces a significant improvement in PEG-fusion success compared to CTL. Further, MEL produces significantly (p<0.001) larger peak CAP amplitudes conducted through the lesion site following PEG-fusion compared to CTL or any other experimental compound in vitro. Additionally, MEL significantly (p<0.025) increases the ability to PEG-fuse sciatic axons in vivo, compared to CTL. Finally, PEG-fusion success in vivo is significantly (p<0.01) greater in calcium-free CTL (CTL-Ca) compared to CTL.