Abstract.In this paper an attempt is made to verify the hypothesis of the role of geomagnetic disturbances as a factor in determining the intensity of traveling ionospheric disturbances (TIDs). To improve the statistical validity of the data, we have used the method involving a global spatial averaging of disturbance spectra of the total electron content (TEC). To characterize the TID intensity quantitatively, we suggest that a new global index of the degree of disturbance should be used, which is equal to the mean value of the rms variations in TEC within the selected range of spectral periods (of 20-60 min, in the present case). The analysis has been made for a set of 100 to 300 GPS stations for 10 days with a different level of geomagnetic activity (Dst from 0 to −350 nT; the Kp index from 3 to 9).It was found that power spectra of daytime TEC variations in the range of 20-60 min periods under quiet conditions have a power-law form with the slope index k = −2.5. With an increase in the level of magnetic disturbance, there is an increase in the total intensity of TIDs, with a concurrent kink of the spectrum caused by an increase in oscillation intensity in the range of 20-60 min. The TEC variation amplitude is found to be smaller at night than during the daytime, and the spectrum decreases in slope, which is indicative of a disproportionate increase in the amplitude of the small-scale part of the spectrum.It was found that an increase in the level of geomagnetic activity is accompanied by an increase in the total intensity of TEC; however, it does not correlate with the absolute level of Dst, but rather with the value of the time derivative of Dst (a maximum correlation coefficient reaches −0.94). The delay of the TID response of the order of 2 hours is consistent with the view that TIDs are generated in auroral regions, and propagate equatorward with the velocity of about 300-400 m/s.Correspondence to: E. L. Afraimovich (afra@iszf.irk.ru)