Changes in postsynaptic Ca2+ levels are essential for alterations in synaptic strength. At hippocampal CA3-to-CA1 synapses, the Ca2+ elevations required for LTP induction are typically mediated by NMDA receptor (NMDAR) channels but a contribution of NMDAR-independent Ca2+ sources has been implicated. Here, we tested the sensitivity of different protocols modifying synaptic strength to reduced NMDAR-mediated Ca2+ influx by employing mice genetically programmed to express in forebrain principal neurons an NR1 form that curtails Ca2+ permeability. Reduced NMDAR-mediated Ca2+ influx did not facilitate synaptic depression in CA1 neurons of these genetically modified mice. However, we observed that LTP could not be induced by pairing low frequency synaptic stimulation (LFS pairing) with postsynaptic depolarization, a protocol that induced robust LTP in wild-type mice. By contrast to LFS pairing, similar LTP levels were generated in both genotypes when postsynaptic depolarization was paired with high frequency synaptic stimulation (HFS). This indicates that the postsynaptic Ca2+ elevation also reached threshold during HFS in the mutant, probably due to summation of NMDAR-mediated Ca2+ influx. However, only in wild-type mice did repeated HFS further enhance LTP. All tested forms of LTP were blocked by the NMDAR antagonist D-AP5. Collectively, our results indicate that only NMDAR-dependent Ca2+ sources (NMDARs and Ca2+-dependent Ca2+ release from intracellular stores) mediate LFS pairing-evoked LTP. Moreover, LTP induced by the first HFS stimulus train required lower Ca2+ levels than the additional LTP obtained by repeated trains.