AbstractIrrigation modulates the water cycle by making water available for plants, increasing transpiration and atmospheric humidity, while decreasing temperatures due to the energy that is needed for evaporation. Irrigation is usually not included in atmospheric reanalysis systems, but moisture can be added to the soil due to data assimilation. This paper compares these soil moisture additions to the irrigation patterns. In the ERA‐interim atmospheric reanalysis, 2 m temperature observations are assimilated. A mismatch between modeled and observed temperatures is corrected by adding or removing moisture from the soil. These corrections show a clear pattern of mean soil moisture additions in many areas. To determine the cause of these increments, the spatial and temporal patterns of these soil moisture increments are compared to irrigation water demand and precipitation bias. In irrigated areas, the annual means and cycles of soil moisture increments correlate well with irrigation, and less with precipitation bias. Therefore, in irrigated areas, the soil moisture increments are more likely caused by irrigation than by the precipitation bias. In nonirrigated areas, a weak statistical relation between soil moisture increments and precipitation bias is present. Irrigation is currently not included in reanalysis systems. However, as irrigation indirectly influences the water balance in atmospheric reanalysis systems, we recommend to include this process in reanalysis models. Moreover, the influence of irrigation on the local and regional atmosphere should be taken into account when interpreting atmospheric data over strongly irrigated areas.