Delayed treatment with CHPG improves functional recovery after TBI. (A) Motor coordination was assessed using a beam walk test in TBI mice at one, two and three months post-injury. All injured mice performed poorly on this task prior to treatment, with no significant differences between treatment groups. At two and three months post-injury, CHPG-treated TBI mice showed a significant increase in the number of correct hind limb placements when compared to vehicle-treated mice (*P < 0.05 versus vehicle). Co-administration of MTEP with CHPG blocked CHPG-mediated improvements such that this group performed similar to vehicle-treated TBI mice (n = 7 (vehicle), 7 (CHPG), 4 (MTEP + CHPG)). (B) Spatial learning and memory was assessed using a one-day MWM protocol performed three and a half months post-TBI. The latency to find the submerged hidden platform in trial 4 was significantly increased in vehicle- and MTEP + CHPG-treated TBI mice when compared to naïve controls (P <0.05 versus naïve). CHPG treatment significantly reduced the latency when compared to vehicle-treated TBI mice (P < 0.05 versus vehicle) (n = 4 (naïve), 8 (vehicle), 8 (CHPG), 8 (MTEP + CHPG)). (C) One day after MWM mice performed a probe trial in which the platform was removed and the time to enter the correct quadrant where the platform had been was measured. Vehicle-treated TBI mice had significantly increased latency to first entry into target quadrant compared to naïve controls (P < 0.05 versus naïve). CHPG-treated TBI mice had reduced retention memory deficits and had significantly reduced latency to first entry (P < 0.01 versus vehicle). (D) There were no significant differences in swim speeds between any of the treatment groups throughout the trials. (E) The escape strategy in the MWM test was analyzed, and the percent composition for each strategy (spatial, systematic and looping) demonstrated that vehicle- and MTEP + CHPG-treated TBI mice had diminished spatial and systematic strategies and increased looping strategies compared to naïve or CHPG-treated TBI mice. Statistical analyses were by repeated measures one-way ANOVA with Student-Newman-Keuls post-hoc corrections in (A) and (B), one-way ANOVA with Student-Newman-Keuls post-hoc corrections in (C) chi-square analysis in (E). Bars represent mean ± SEM.