The purpose of this study was to explore morphologic, functional, and behavioral effects of rough (sandblasted-large grid/acid-etched (SLA) and TiO2 blasted), mechanically polished, and mechanochemically polished titanium implant surfaces on nerves. The compound action potentials (cAPs) of sciatic nerves of sacrificed Wistar rats (n=10) were quantified at the in vitro level, while contacting disk-shaped test specimens. The test specimens were also implanted directly on the sciatic nerves of another group of animals (n=33), hot-plate tests were undertaken for 10 consecutive days, and then the animals were sacrificed. Quantification of signal transduction speeds and cAPs of the nerves of these animals were undertaken at the in vitro level. Finally, the nerves were processed for histologic analysis. The signal transduction speeds and duration of cAPs of all groups were similar (P>0.05), whereas the amplitudes of cAPs of nerves contacting SLA implants were higher than those of TiO2 blasted and mechanochemically polished surfaces (P<0.05). Response latencies of nerves contacting mechanically polished specimens were slightly higher than the other groups (P>0.05). Histologic evaluations did not reveal any signs of adverse tissue response adjacent to specimens tested. Rough and polished titanium implant surfaces lead to similar neural response in vivo and in vitro that fall into physiologic limits.