Image retrieval was a matching process between feature of query image and feature of target image. Image retrieval was a very interesting project because many institutions such as artistic gallery and police used it. This paper explained about image retrieval method that was used to retrieve image from a collection of images. The method was sorted wavelet histogram. The method was a result of development of previous image retrieval which implements histogram to wavelet transforms coefficients for feature extraction. The retrieval application just runs in the computer when the test was carried out. The result showed that this method gave good performance which the average of succeed rate was 100% for original image, 99,83% for the image of Blur, 90,25% for edge image and 73,57% for the 50% salt & paper noise image and the querying time is about 0.2 second.