The main purpose of this study was to show potential uses of very high resolution (VHR) image in an urban analysis of Bukavu in the Democratic Republic of Congo. As for many of the Third-World cities, Bukavu grew up during the last decades and available topographical information on Bukavu dates back to the middle of the twentieth century. This lack of updated information can be compensated by an appropriate use of VHR images. In this study, IKONOS image recorded on the 14 th February 2001 was used. Image registration are highly dependant on accurate Digital Elevation Models (DEM), these last ones are also useful in urban analyses. The range between the minimum and maximum altitudes observed in the image was more than 500 meters. Furthermore, the viewing inclination angle is more than 28 degrees. In this case, orthorectification is mandatory for correcting relief displacements. Nevertheless, the lack of good Ground Control Points (GPS’s) on the old topographic maps and the failure to collect field verification data in Bukavu explains the remaining global 2D RMSE of 10 meters. For easier image interpretation, multispectral (4m) and panchromatic (1m) images were fused together by means of the LMVM algorithm. Depending on the object, the Computer Aided Photo Interpretation (CAPI) uses or d oes not use the near infrared information (true or false colour composite). For a more detailed interpretation about the city morphology we draped the 1m multispectral fused image over the 1m resolution DEM grid. The Built-up Area Index (BAI) computed on the urban mask obtained by CAPI and classification of the vegetation, was compared with the 1954 situation interpreted from topographic maps. The present city shows higher BAI values and in the same time the centre of the city has clearly shifted southward. Statistical analyses are also done on built-up versus slope data. The lack of good GCP’s and the use of DEM produced inaccurate orthorectification which was not adequate for topographical features extraction. Nevertheless the present extensions of Bukavu built-up areas are more than the double of those observed at the beginning of the last half century. New constructions are located on steeper slopes where landslides are frequents. More features could be extracted from the Ikonos image if good GPS measurements were made and if verification by the city authorities was possible. Some practical applications of this study could involve determining better location of new Bukavu extensions.