We present a method to integrate case-based reasoning (CBR) with object-oriented image classification to classify SPOT images. Images were first segmented into discrete objects at multiple scales. CBR was then used to classify these objects by comparing their geometric shapes, spectral characteristics, and textural measurements with those of the past objects prepared from archived SPOT images and land use data. Once enough past objects were accumulated, this method was able to successfully classify image objects with promising results as demonstrated by a case study of aquaculture land use mapping in coastal areas of Guangdong province, China.