Long-term potentiation of hippocampal excitatory synapses is often accompanied by an increase in the probability of spiking to an EPSP of fixed strength (E-S potentiation). We used computer simulations of a CA1 pyramidal neuron to test the plausibility of the hypothesis that E-S potentiation is caused by changes in dendritic excitability. These changes were simulated by adding "hot spots" of noninactivating voltage-sensitive Ca2+ conductance to various dendritic compartments. This typically caused spiking in response to previously subthreshold synaptic inputs. The magnitude of the simulated E-S potentiation depended on the passive electrical properties of the cell, the excitability of the soma, and the relative locations on the dendrites of the synaptic inputs and hot spots. The specificity of the simulated E-S potentiation was quantified by colocalizing the hot spots with a subset (40 of 80) of the synaptic contacts, denoted "tetanized," and then comparing the effects of the hot spots on these and the remaining (untetanized) synaptic contacts. The simulated E-S potentiation tended to be specific to the tetanized input if the untetanized contacts were, on average, electrically closer to the soma than the tetanized contacts. Specificity was also high if the tetanized and untetanized contacts were segregated to different primary dendrites. The results also predict, however, that E-S potentiation by this mechanism will appear to be nonspecific (heterosynaptic) if the synapses of the untetanized input are sufficiently far from the soma relative to the tetanized synapses. Experimental confirmation of this prediction would support the hypothesis that changes in postsynaptic excitability can contribute to hippocampal E-S potentiation.