Errors introduced into aeromagnetic survey results because of geomagnetic time variations have been analyzed in two ways. In the first, statistical properties of average magnetic disturbance have been calculated for five locations well distributed in latitude in Canada. The autocorrelation functions of vertical magnetic field fluctuations have been used to estimate errors in aeromagnetic surveying as a function of the length of the base loop and the length of the traverse lines or profiles, for five latitudes. In practical applications, the r.m.s. errors are proportional to the square root of elapsed time. These results can be explained in terms of the shape of the autocorrelation functions, and models of disturbance which might produce them are discussed. In the second part experimental results are presented of a comparison of total field intensity fluctuations measured with recording proton precession magnetometers at two pairs of stations, one pair 26 miles apart and the other 94 miles apart. The measurements were taken in Western Canada south of the auroral zone. The r.m.s. differences in the time variations have been determined quantitatively for each hour for both pairs of stations; statistically, it was found that these differences were proportional to the r.m.s. level of magnetic activity and, in this region, to the separation of the stations. The frequency distributions of the r.m.s. differences between the two pairs of stations for the month of September 1959 have been determined. A comparison of the indirect and direct methods of correction indicates that scientifically, in this region, there is little to choose between them in terms of accuracy. Operational considerations are mentioned.