The work is describing an image database consisting of digitized mammograms and all range of available information about patients’ past history and clinical reports. Also, information about mammography equipment used to obtain the original films was included in the database. A Windows-based interface was created so that the database can be used as a supporting tool for tests of procedures of segmentation and detection of structures such as microcalcifications, nodules and masses. Images from exams performed in 5 different hospitals were used, which provide a good diversity regarding mammography equipments and image classes. Clinical information about the patients and radiological reports was also included, pointing out structures of clincial interest. The query to the database can be done by choosing hospital, patient age and structure present in the image. A routine to search pre-selected regions of interest is also included. Two laser scanners (Lumiscan 50 and Lumiscan 75) were used to digitize the mammograms. They can digitize images with 12-bit of contrast resolution with 150 m and 75 m of spatial resolution respectively. Currently about 3500 images corresponding to more than 600 patients are available in the database. All of them has provided radiological reports as well as biopsies reports when pertinent. Database has been used currently to select images mainly for testing processing image techniques developed in our research workgroup regarding our computer-aided diagnosis scheme in mammography.