The gating of ion channels has widely been modeled by assuming that the transitions between open and closed states are a memoryless process. Nevertheless, analysis of records of unitary current events suggests that the kinetic process presents long lags (antipersistent correlation). Here, using the patch-voltage clamp technique and the rescaled range method, activity of single-channel delayed rectifier K(+) channels was studied. The experiment result showed that reversal potential was -73.3 mV in cell-attached mode. For the sequences of alternating open and shut time intervals, the Hurst coefficients were calculated for four different pipette potentials in rat dorsal root ganglion neurons. H=0.34169+/-0.00672 (n=4) for V=-30 mV; H=0.34632+/-0.0142 (n=3) for V=-40 mV; H=0.39237+/-0.0113 (n=4) for V=-50 mV; H=0.3954+/-0.0012 (n=4) for V=-60 mV. When the Hurst method was applied to the results from a simulated four-state Markovian model, it showed that it had different experimental data H coefficient, the distribution of the data values had no correlations between them, in particular, H=0.2531+/-0.00403 (n=50) for V=-40 mV. This indicates that open-dwell times and closed-dwell times are long lag (namely, antipersistent correlation) and do not change with the pipette potential applied to the patch.