Synaptic plasticity has long been considered as a cellular basis for learning and memory. The most prominent forms of such plasticity are the homosynaptic long-term potentiation (LTP) and long-term depression (LTD) in the hippocampus. The aim of the present study has been to advance the understanding of the mechanisms involved in the expression of hippocampal LTP/LTD. Experiments were performed on hippocampal slices from 12-18 day old Sprague-Dawley rats. Standard electrophysiological techniques for extracellular recording were used. The results showed that the homosynaptic LTD was associated with equal relative changes of AMPA and NMDA responses. LTP, on the other hand, occurred as an early (<30-60 min) change of predominantly the AMPA response, and a later one associated with equal increases of AMPA and NMDA responses. An adenosine agonist N6-cyclohexyladenosine (CHA) mimicked the parallel change of AMPA and NMDA responses during LTD. Furthermore, the effect of LTD on the waveforms of the isolated NMDA and AMPA responses was found compatible to the action of several drugs attenuating synaptic transmission by presynaptic, but not postsynaptic, mechanisms. However, experiments using the activity dependent NMDA channel blocker (+)-MK-801 hydrogen maleate (MK-801) failed in demonstrating a decrease in transmitter release during LTD. There was also a lack of interaction between LTD and presynaptic changes such as CHA-induced depression and paired-pulse facilitation. It was also found that LTD and LTP were able to repeatedly reverse each other, with respect to both AMPA and NMDA responses. It was further demonstrated that not only LTD, induced by special induction paradigms, could reverse LTP but also the test stimulation by itself facilitated the early decay of LTP. The results suggest that LTD and (later part of) LTP involve a common synaptic modification, although in different directions. A possible explanation of our data is that this common modification takes place as a co-ordinated pre- and postsynaptic change. In principle, this could occur by alterations of existing synapses. Alternatively, elimination and formation of (functional) synapses could be involved, perhaps by functionally silencing synapses or bringing such synapses into play.