To investigate the after-effects of 25-Hz repetitive transcranial magnetic stimulation (rTMS) at 60, 100, and 120% resting motor threshold (rMT) on long-term potentiation (LTP) in the rat hippocampus, to clarify the intensity dependence of rTMS, and to determine whether it simultaneously affects learning and memory ability. Five rats were randomly selected from 70 male Wistar rats, and evoked rMT potentials were recorded in response to magnetic stimulation. The remaining 65 rats were randomly assigned to five groups (n = 13), including sham rTMS, 1 Hz 100% rMT, and 25 Hz rTMS groups with 3 subgroups of 60% rMT, 100% rMT, and 120% rMT. Five rats in each group were anesthetized and induced by a priming TMS-test design for population spike (PS) response of the perforant path-dentate gyrus in the hippocampus; the remaining eight rats in each group were evaluated for object recognition memory in the novel object recognition (NOR) task after the different rTMS protocols. Forty-five percent (approximately 1.03 T) of the magnetic stimulator output was confirmed as rMT in the biceps femoris muscle. The PS ratio was ranked as follows: 25 Hz 100% rMT (267.78 ± 25.71%) > sham rTMS (182 ± 9.4%) >1 Hz 100% rMT (102.69 ± 6.64%) > 25 Hz 120% rMT (98 ± 11.3%) > 25 Hz 60% rMT (36 ± 8.5%). Significant differences were observed between the groups, except for the difference between the 25 Hz 120% rMT and the 1 Hz 100% rMT groups (p = 0.446). LTP was successfully induced over the 60-min recording period only in the sham rTMS and 25 Hz 100% rMT groups. Moreover, these two groups spent more time exploring a novel object than a familiar object during the NOR task (p < 0.001), suggesting long-term recognition memory retention. In the between-group analysis of the discrimination index, the following ranking was observed: 25 Hz 100% rMT (0.812 ± 0.158) > sham rTMS (0.653 ± 0.111) > 25 Hz 120% rMT (0.583 ± 0.216) >1 Hz 100% rMT (0.581 ± 0.145) > 25 Hz 60% rMT (0.532 ± 0.220). The after-effect of 25-Hz rTMS was dependent on stimulus intensity and provided an inverted (V-shaped) bidirectional modulation on hippocampal plasticity that involved two forms of metaplasticity. Furthermore, the effects on the recognition memory ability were positively correlated with those on LTP induction in the hippocampus in vivo.