Understanding nitrogen and phosphorus leaching rates is critical for controlling non‐point source pollution. As of now, no studies have dealt with the spectrums of N or P leaching loss, a new concept is suggested in the present study, which can be used to assess the risks of N or P leaching loss. The spectrums of N or P leaching loss consist of the highest leaching loss (HLLr), mean leaching loss (MLLr), lowest leaching loss (LLLr), lowest leaching loss at an upper limit (LLLU), and lowest leaching at a lower limit (LLLL), and annual average of the lowest leaching loss (AALLL) and annual average of different‐level leaching loss (AADLL). These indices are tested with leaching experiments with distilled water, lysimeters, and PVC‐pipe method based on Darcy's law on seven land‐use types in the Dachong karst watershed in China. LLLL is similar among all land‐use types, except vegetable land. Four types of forests show few differences in LLLU and LLLL. HLLr, MLLr and LLLr, and AALLL and AADLL in the croplands are clearly greater than those in the forests. The HLLr of total nitrogen in vegetable land is highest (16.08 mg L−1) but the HLLr of nitrate nitrogen, ammonium nitrogen, and total phosphorus in paddy field are highest (11.01, 1.39, and 0.3 mg L−1, respectively). LLLr, MLLr, and HLLr in the same type of land use are one to three times higher than LLLU in N leaching loss. The spectrums intuitively show differences in N and P leaching losses among land‐use types in the karst watershed in China.