CA1 pyramidal cell is modeled by a linked series of passive compartments representing the soma and different parts of the dendritic tree. Intracellular postsynaptic potentials are simulated by conductance changes at one or more compartments. By assuming an infinite homogeneous extracellular medium and a particular geometrical arrangement of pyramidal cells, field potential profiles are generated from the current source-sinks of the compartments. The pyramidal cells are driven at the theta (theta)-frequency at different sites of the dendritic tree in order to simulate external driving of hippocampus by the septal cells. Inhibitory or excitatory driving at different sites gives extracellular dipole fields of different null zones and maxima. Phase reversal (180 degrees) of a dipole field generated by synchronous synaptic currents is completed within a depth of 150 micron. By driving two spatially distinct but overlapping dipole fields slightly phase-shifted (30-90 degrees) from each other, the resultant field shows a gradual phase shift of 180 degrees in over 400 micron depth and no (stationary) null zones. The latter field correspond to the theta-profiles seen in the freely moving rat. Somatic inhibition is proposed to be the synaptic process generating the theta-field potentials (named dipole I) in the urethananesthetized or curarized rat. Dipole I has amplitude maxima at the basal dendritic and the distal apical dendritic layers, with a distinct null zone and phase reversal at the apical side of the CA1 pyramidal cell layer. Rhythmic distal dendritic excitation, time-delayed to somatic inhibition, is proposed to be the additional dipole (dipole II) found in freely moving rats. The combination of dipoles I and II, phase-shifted from each other, causes the gradual theta-field phase shift. Experimental studies indicate that dipole I is atropine-sensitive and probably driven by a cholinergic septohippocampal input, whereas dipole II is atropine-resistant and may come from a pathway through both the septum and the entorhinal cortex. Variations of the phase profiles of the theta-field in freely moving rats by administration of anesthetic and cholinergic drugs and by normal changes in theta-frequency could be accounted for by the proposed model. Changes of the intracellular membrane potential, cellular firing rate, and evoked excitability at different phases of the theta-rhythm in anesthetized and freely moving rats can be predicted from the model, and they are in general agreement with the extant literature. In conclusion, theta-field is generated by a rhythmic somatic inhibition phase-shifted with a distal apical-dendritic excitation.(ABSTRACT TRUNCATED AT 400 WORDS)