A series of experiments has been conducted with probes made from silicon, glass, tungsten and polyimide within a developed brain phantom, and the insertion behavior, forces and dimpling are compared to in vitro and in vivo models. This allows the choice of proper insertion parameters and probe structure to reach a compromise between needle stability and tissue trauma as a result of insertion. According to the performed experiments, the reduced interfacial area between the needle tip and the brain will result in reduced insertion force. High insertion speed (100 mm/min) reduces the dimpling but not the penetration force necessarily. In vivo insertion and retraction of the fragile probes made from silicon is possible without pia and/or dura removal.