In this study, the accuracy of steady and unsteady methods in predicting the required length of air-earth heat exchangers have been compared for cooling application. Also, the impact of different conditions for inlet air temperature (constant, periodic and actual temperature) on unsteady methods have been studied. Results indicated that, steady model cannot provide correct estimation for the required length of channel. In addition, using unsteady method with constant inlet air temperature overestimates the required length of the channel, which can lead to economic and implementation challenges for project. However, using unsteady method with periodic inlet air temperature the required length of the channel can be estimated with acceptable accuracy compared with real situation (unsteady model with actual inlet air temperature).Results showed that after a month of the launch of the system, its thermal behavior reaches to steady state level. During this period effectiveness of the system drops significantly (nearly 50%). Nearly half of the total performance drop of the system occurs during the first working day. It means that the long-term behavior of the system can be determined by monitoring of its behavior in the first day of operation. Results also demonstrated that using common methods of designing usually brings a significant error.