Field-potential recordings with macroelectrodes, and extra- and intracellular potentials with micropipettes were used to determine the influence of spontaneous field potentials on the activity of neocortical neurons during seizures. In vivo experiments were carried out in cats under anesthesia. Strong negative field fluctuations of up to 20 mV were associated with electroencephalogram “spikes” during spontaneously occurring paroxysmal activities. During paroxysmal events, action potentials displayed an unexpected behavior: a more hyperpolarized firing threshold and smaller amplitude than during normal activity, as determined with intracellular recordings referenced to a distant ground. Considering the transmembrane potential (the difference between extra- and intracellular potential) qualified this observation: firing threshold determined from the transmembrane potential did not decrease, and smaller action-potential amplitude was associated with depolarized firing threshold. The hyperpolarization of intracellular firing threshold was thus related to the field potentials. Similar field-potential effects on neuronal activities were observed when the paroxysmal events included very fast oscillations or ripples (80–200 Hz) that represent rapid fluctuations of field potentials (up to 5 mV in <5 ms). Neuronal firing was phase-locked to those oscillations. These results demonstrate that: ( a) strong spontaneous field potentials influence neuronal behavior, and thus play an active role during paroxysmal activities; and ( b) transmembrane potentials have to be used to accurately describe the behavior of neurons in conditions in which field potentials fluctuate strongly. Since neuronal activity is presumably the main generator of field potentials, and in return these potentials may increase neuronal excitability, we propose that this constitutes a positive feedback loop that is involved in the development and spread of paroxysmal activities, and that a similar feedback loop is involved in the generation of neocortical ripples. We propose a mechanism for seizure initiation involving these phenomena.