Abstract: Current and historical land cover / land use information is an important contribution to many urban and watershed planning activities. This information is invaluable for those investigating the relationship between changes in human activity and changes in water quality. Understanding past activities will assist in the planning of future land development. The main objective of this study was to use medium resolution (Landsat) satellite imagery to produce an accurate historical land use / land cover map along the Lake Erie coastline, focusing on areas located in the rural / urban fringe. This is traditionally a difficult task due to the heterogeneous nature of developed areas. A test area was chosen in the Cleveland, Ohio area because this was a good representation of the rural / urban fringe. The satellite imagery was geometrically and radiometrically corrected to minimize errors created by seasonal and atmospheric variations between images. When initial attempts using traditional pixel-based classification techniques failed to create satisfactory land use / land cover results, an object -- based mapping technique was used. To further improve the quality of the resulting land cover maps, an extra layer -- NDVI difference layers of a leaf on and leaf off image -- was used in the classification process. Also, standard deviation values of Landsat image bands for image objects were added to the classification process to increase separation between land cover classes. The final land cover / land use classification map was produced with an 89.8% overall accuracy. Once land cover maps were created for the study area for each time period a change detection analysis was performed. It was found that the total area of high density urban, forest, industrial, water and exurban or rural land cover type stays relatively constant. The suburban and low density urban areas increase from 1974 to 1994. The suburban land area decreases between 1994 and 2004. Between 1974 and 1994 the predominant land cover change in the study area was from agriculture to suburban and low density urban, which explains both the increase in area for low density urban and suburban and the decrease in area for agriculture. Then between 1994 and 2004 a major land cover change was seen from suburban land cover to low density urban. This would explain the decrease in total area for suburban land cover for this time period.