The present application discloses a speaker box. The speaker box includes an upper cover, a lower cover forming an accommodation space with the upper cover, and a speaker received in the accommodation space. The speaker includes a frame and a diaphragm positioned by the frame. The upper cover forms a supporting wall extending toward the lower cover for fixing the speaker. The speaker box further includes an elastic cushion arranged between the frame and the supporting wall. A method for assembling the speaker box is also provided. Due to the elastic cushion, the consistency of the speaker is ensured, and the acoustic performance of the speaker box is accordingly improved.