Due to various challenging conditions during video recording, such as camera shake and out-of-focus issues, video deblurring remains a difficult problem. To address this, we propose the Spatial-Temporal Frequency domain Fusion network (STFFNet) and improve the network from three key aspects. Firstly, we introduce the Encoder-Decoder idea to create a novel backbone to combine global and local features effectively. Secondly, a new feature fusion module that focuses on the differences between frames is proposed to help better deblur the current frame. Finally, STFFNet introduces a Frequency Domain Converter (FDC) to transform the image information from the spatial domain to the frequency domain, enhancing image restoration by narrowing the gap between the deblurring and ground truth images in the frequency domain. Experimental results demonstrate that the proposed method achieves state-of-the-art deblurring performance on benchmark datasets. The code is available at: https://github.com/Paige-Norton/STFFNet .