Due to the disconnection between scientific fertilization guidance and the agricultural production, excessive application of phosphorus fertilizer is widespread, which results in the accumulation of large amounts of phosphorus in soil. Accurate evaluation of phosphorus availability in soil is the basis for scientific management of soil phosphorus. Diffusive gradients in thin films technology (DGT) can more accurately reflect the bioavailability of elements in soils than traditional methods. In this study, 60 representative soil samples from the Loess Plateau were used to evaluate the status of soil phosphorus pool and its availability in the Loess Plateau under the condition of imbalanced fertilizer input for long period, and traditional Olsen and the DGT methods were used. The relationships between DGT-P, and Total-P and Olsen-P were also studied. Path Coefficient Analysis was used to explore the relationship between soil phosphorus availability and relevant soil physical and chemical indicators. The results showed that the distribution of soil phosphorus was uneven in the Loess Plateau. Phosphorus deficiency and phosphorus enrichment coexisted, specifically, phosphorus rich samples were mainly from farmland. Soil samples with phosphorus activation coefficient (PAC) less than 2.0% accounted for 76.67% of the total samples, and most soil samples had low phosphorus availability. Because phosphorus input in farmland was too high, activating the unavailable phosphorus in soil was the main target of scientific management of phosphorus in the Loess Plateau. Soil DGT-P was significantly correlated with total-P and Olsen-P (P<0.01). Clay content had the greatest direct effect on soil phosphorus availability compared with other factors, and its effect on Olsen-PAC was positive, and the effect on DGT-PAC was negative. The higher the soil clay content, the larger the available phosphorus pool in soil, but the less available phosphorus could be taken up by plants.