Abstract: Background Gully erosion is the most severe form of land degradation in the black soil region of Northeast China, posing a critical threat to agricultural production and ecological security. Methods This study systematically summarizes research advancements on the current status, development characteristics, spatiotemporal evolution and its influencing factors, key processes and mechanisms of gully erosion, as well as gully erosion models, prevention and control technologies, and management models by literature data extraction and analysis, team research data and field investigation. Results Gully exhibits uneven spatial distribution in the region, with 3/4 of all gullies concentrated within 1/3 of the cropland, and gully erosion accounting for over 65% of total soil loss. Seasonal variations in gully erosion are pronounced with headcut retreat being dominated by on-wall flow and soil fragment detachment during the snowmelt period and a "scouring-collapse-headcut retreat" cycle prevailing in the rainy season. Freeze-thaw cycles promote the formation of tensile cracks, which diminish during the rainy season and contribute to 74%–82% of gully bank expansion erosion. The headcut retreat, bank expansion, and incision rates of individual gully are 6.42 m/a, 0.35 m/a, and 0.12 m/a, respectively. The average rates of gully number increase, gully density, and destroyed-land area are 3 gullies/(100 km²·a), 11.56 m/(km²·a), and 209.86 m²/(km²·a), respectively, with significant spatial scale effects. The high-intensive reclamation of forest and grasslands is the primary driver of gully formation, and returning farmland to forests/grasses can mitigate erosion but its short-term efficacy is limited. Gully erosion is most severe in areas with 2-8° slopes, sunny exposures, 500-1000 m slope lengths, and 240-300 m elevations. Climate, topography, soil profile structure and its underlying geological conditions, ridge tillage methods, and their interactions significantly affect gully development process. Cross-slope ridge tillage exacerbates headcut retreat during heavy rainfall due to a "channel system effect". Gully volume models, topographic threshold models, and sensitivity evaluation models had a substantial regional differentiation. Deep learning approaches integrating label correction, lightweight networks, and multi-source data fusion effectively identify gullies. Four major gully control models, integrating engineering, vegetation, and land reclamation, have been developed and used popularly in the black soil region. Future research should prioritize precision monitoring of gully erosion, multi-factor coupled driving mechanisms, construction of dynamic and process models, intelligent gully identification, and development of evaluation systems for governance benefits. Conclusions These efforts will provide scientific support for advancing fundamental research, guiding management decisions and planning, and optimizing engineering implementation related to gully erosion in the black soil region of Northeast China.