Spatial patterns of soil faunal communities at a regional scale are important foundations for the maintenance and protection of belowground biodiversity. Although many studies have addressed the spatial patterns of soil animal communities in different habitats, few have focused on the spatial patterns along latitudinal gradients. In this study, we determined the composition, horizontal structure, vertical structure, and diversity characteristics of soil macrofaunal communities in farmland systems along a latitudinal gradient. We selected three representative mountain areas along the latitudinal gradient:the Maoer Mountains (Maoer Mountain ecosystem station, 45°20'-45°25'N, 127°30'-127°34'E), the Xiaoxing'an Mountains (Dailing,47°10'-47°14'N, 128°53'-128°55'E), and the Da Hinggan Mountains (Tahe, 52°09'-53°23'N,123°19'-125°48'E). In each study area, three farmlands at least 50 m apart with similar conditions were randomly selected. In each farmland, four plots (5 m×5 m) were randomly delineated. In each plot, four soil profiles (50 cm×50 cm) were randomly collected and divided equally into three 5 cm strata to survey the soil macrofaunal communities. The hand picking method of separation was used to determine the soil macrofaunal communities. Fieldwork was conducted in August 2011. Using bivariate related analysis and canonical correspondence analysis, we assessed the relationships between soil animal communities and environmental factors. In total, 2339 individuals from 35 groups of soil animals were collected, which belonged to 2 phyla, 6 classes, 14 orders, and 35 families. Enchytraeidae, Lumbricidae, Carabidae, and Staphylinidae were dominant families, accounting for 58.84% of the total number of individuals. Formicidae, Araneidae, and Geophilomorpha were common groups, accounting for 32.79% of the total number of individuals. The results showed that:(1) In terms of horizontal distribution, individual density and richness (i.e. numbers of groups) at sites followed the order Maoer Mountain > Dailing > Tahe. Results of the one-way ANOVA revealed there were no significant differences in individual density or richness in the soil macrofaunal communities at different latitudes. The Shannon-Wiener diversity index (H) and Pielou evenness index (E) for sites followed the order Maoer Mountain > Dailing > Tahe. The Simpson dominance index (C) of Tahe and Dailing were higher than that of the Maoer Mountains, and the Margalef richness index (D) was greatest in Tahe, followed by the Maoer Mountains and Dailing. (2) In terms of vertical distribution, no significant differences were detected among the three different latitudinal areas for individual density or richness at the same vertical level (P < 0.05). Individual density and richness in soil macrofaunal communities decreased with increasing soil depth, except for soil macrofaunal communities from the Maoer Mountains at 5-10 cm, indicating an obvious surface aggregation. (3) Regarding the relationships between soil macrofaunal communities and soil environmental factors, the bivariate correlation analysis showed there were no significant correlations among richness, individual densities, diversity indices, evenness indices, richness indices, pH values, organic matter, total nitrogen, total phosphorus, and available phosphorus. Results of canonical correspondence analysis further indicated that the dominant and common groups were highly adaptable to environmental factors and the dominant and common groups were widely distributed in the three latitudinal areas. This study showed that the numbers of groups, individual densities, and diversity indices of the soil macrofaunal communities from the farmland ecosystem changed along a latitudinal gradient, but no significant changes were found in this study. Different soil animal groups at different latitudes are affected by environment variables. At the local scale, the influence of environmental factors on soil macrofauna should not be ignored. The results of the study create a foundation for researching spatial patterns and mechanisms of biodiversity maintenance at regional scales.