The geomagnetic field components are periodically measured at repeat stations. The main objective of the repeat stations is to provide data for tracing the secular variation of the geomagnetic field components. Secular variation at the repeat station is generally different from that at geomagnetic observatory used in the data reduction. The effect of the secular variation differences on geomagnetic data reduction was estimated for the regions of Europe, North America (below 60°N) and Australia, respectively, during the period of 2000-2010. These estimations were obtained by using the monthly mean values of north, east and vertical components of geomagnetic field, recorded at geomagnetic observatories. The effects were calculated by using observatories pairs, with distances from 350 km (in Europe) to 3100 km (in North America and Australia). The maximal effects were found to be the smallest in east component in Europe and North America, and vertical component in Australia; the effects increase with time from a central reduction epoch and they are not constant during mentioned eleven years; they were less than 1 nT only in Europe, for distances between the observatories up to 1000 km in all three components and for periods spanning ±1 month from a central epoch. It was found that their year to year variability is mostly due to the non-eliminated external field residuals in the observatories monthly means; their effect is up to 3 nT for ±3 months from a central epoch. Further, the real effects were compared to those modeled by IGRF-12 model. The maximal differences between the real and the modeled values are below 4.5 nT in all three components, for ±3 months from a central epoch.