Latency to platform of young adult and aged rats both at control and surgery groups decreased with the increase of training days. No significant difference was detected between control and surgery groups of young adult and aged rats by repeated measures of ANOVA (P = 0.202 for aged; P = 0.057 for young adult) (A). During probe test of aged rats, repeated measures of ANOVA showed that the latency for the first entrance of targeted area wasn’t affected by surgery (P = 0.495), measure time (P = 0.067), and the interaction of surgery×measure time (P = 0.984) (D). Yet, percents of time (B) and distance (C) in the targeted area both were significantly affected by surgery (P = 0.045 for time; P = 0.041 for distance), measure time (P = 0.013 for time; P<0.001 for distance), and the interaction of surgery×measure time (P = 0.037 for time; P = 0.013 distance). Percents of time (B) and distance (C) in the targeted area of aged rats 1d (P = 0.006 for time; P<0.001 for distance) and 3d (P = 0.011 for time, P = 0.002 for distance) after surgery all were obviously less than that of 1d before surgery, respectively. For young adult rats, there was no obvious difference at the percent of distance in the targeted area (P = 0.170), the percent of time in the targeted area (P = 0.772) and the latency for the first entrance of targeted area (P = 0.938) between control and surgery groups 1d before surgery (A). Percents of time and distance, and the latency for the first entrance in the targeted area were not affected by surgery (P = 0.639 for time; P = 0.773 for distance; P = 0.983 for latency) and measure time (P = 0.327 for time; P = 0.358 for distance; P = 0.173 for latency) (B, C, D). 1d pre: 1d before surgery; 1d aft, 3d aft and 7d aft: 1d, 3d, 7d after surgery. Data were mean ± SEM. *p<0.05 vs 1d before surgery.