Abstract: Step-down relative humidity (RH) means that the RH is gradually reduced to improve the drying efficiency and quality in hot air drying. The hot air drying with step-down RH has been successfully applied to the yam slices, Dahongpao pepper, and papaya slices drying. This study aims to regulate the step-down RH in the period of dehumidification, in order to improve the efficiency and quality. Three stages of dehumidification were two-, multi-stage dehumidification drying and RH control, according to the material temperature. The internal moisture migration and surface moisture were also selected to represent the changes in RH. Specifically, there was an increase in the surface moisture evaporation, when the RH shared an upward trend. Once the RH showed a downward trend, the internal water moisture increased much more than the surface water evaporation. As such, the RH control regime was as follows, according to the RH changes. In the early drying, the evaporation of the material itself caused the RH rise. The middle drying then reached, after the RH was stable. If the RH shared a downward trend, the dehumidification was closed to improve the RH, in order to reduce the surface water evaporation (E) and improve the internal water migration (D). If the RH was rising, the dehumidification was opened to reduce the RH for the high E. If the moisture evaporation on the surface was not enough to increase the RH, or the temperature of the material approached the drying medium temperature, the drying was transferred to the later stage. At the same time, the dehumidification was also opened to reduce the RH for the high E value. The drying end point was reached when the change rate of RH was less than the critical range. Under a drying temperature of 60 ℃ and air velocity of 3.0 m/s, the drying test of the carrot showed that 0-8 min was the early drying, and the RH change rate was less than 0.5% at 8 min. Afterward, the carrot entered the middle drying. The duration of the RH decreasing trend was gradually shortened in the drying period from 8 to 131 min. Whereas, there was the extended in the duration of RH rising trend. Within 137-142 min, there was no rising trend of RH in the period of late drying. At 370 min, the RH change rate was less than 1%/min, indicating that the drying process ended. The RH control mode was used to increase the D and decrease the E value. Consequently, the D and E values remained basically equal within 0.2-2.3 h. Meanwhile, the material temperature presented a step-rising trend correspondingly. Within 0-0.2 h, the accumulation of water (Q) raised rapidly in the material, where the layer of water film was formed to wrap, particularly without the outstanding crust. Within 0.2-2.3 h, the Q value fluctuated up and down at the zero point. The drying rate decreased gradually to produce three zero points in total. The water migrating to the surface was immediately evaporated on the surface without accumulation, indicating the delaying time of crust occurrence to remove a large amount of water. Furthermore, the Q value was gradually less than 0, where the surface of the material produced the outstanding crust to gradually thicken after 2.3 h. After that, the rehydration ratio and shrinkage rate were (4.41±0.02) g/g and (27.32±1.51) %, respectively, at the drying time of 5.6 h. The drying time was shortened by 24.6%, compared with the constant 20%RH. More water migration channels were retained to realize the automatic control of RH. The finding can also provide the theoretical basis and technical support for the RH control during hot air drying of fruits and vegetables.