Abstract: Temperature is one of the most important environmental parameters for fruit or vegetable cold storage. During loading process of fruits or vegetables into cold storage, the new loading goods will influence the temperature stability of goods already settled in the storage room which have reached the storage requirements. Obtaining temperature variation and distribution characteristics in a cold storage during the loading process can provide theoretical reference for optimal design of the cold storage and improve its practical operation and management. To study the factors on the stability of goods temperature during the loading process of fruits and vegetables into cold storage, a combination of CFD (computational fluid dynamics) calculation and experimental study was carried out to further understand the complicated phenomenon of air flow and heat transfer in cold storage. A 3D (three-dimensional) unsteady numerical calculation model of storage environment was developed for a real cold storage with 50 t volume for apples in Xi'an. A ceiling air cooler was equipped in the cold storage. The k-ε turbulent model was considered as CFD model. The apple zone was regarded as porous medium zone. The influences of the spacing of goods, loading temperature and loading quantity on the temperature stability of goods were numerically studied under the condition that the door was closed and the ceiling air cooler was running. The simulation results were validated by experimental results. The experimental system consisted of temperature sensor and temperature illustration cabinet. The temperature sensor was set at ceiling air cooler's air outlet and connected to the temperature illustration cabinet at test room. The temperature was measured and analyzed once every 2 h. In this study, the reliability of the simulation data was verified by the data of the temperature variation within 24 h after the new goods entered the cold storage. The numerical results were in good agreement with the measured results. The non-uniformity coefficient of temperature was introduced to evaluate the temperature stability of the cold storage. Simulation results showed that during the loading process of fruits and vegetables into the cold storage, increasing the spacing of new loading goods could reduce non-uniformity coefficient fluctuation range of temperature. But almost the same value of non-uniformity coefficient of temperature was obtained at the spacing of 0.8 and 0.5 m, which meant that the loading spacing was not the larger the better. The larger loading spacing would increase additional workload for settling apples. Thus, there existed a reasonable small space to ensure the storage quality of goods. Pre-cooling could be helpful to reduce the goods' cooling load. When the loading temperature was decreased from 281.15 to 276.15 K, the maximum value of non-uniformity coefficient of temperature was reduced by 50.3%. The impact of loading quantity on indoor thermal environment was also investigated with 5%, 10% and 15% of storage amount, respectively. When the loading quantity was decreased from 15% to 5%, the maximum cooling time was reduced by 23 h and the maximum value of non-uniformity coefficient of temperature was reduced by 73.7%. The reasonable loading measure should be "small quantity and multiple batches" which can effectively reduce the goods cooling load and the cooling time for the practical operation and management of the cold storage. For the cold storage, the reasonable loading condition is proposed to be 3℃ for loading temperature, 0.5 m for loading spacing and less than 10% for loading quantity.