At our hospital, an average MRI examination comprises 170 images. These examinations were evaluated exclusively on digital reporting systems with two high-resolution monitors for a period of 1 year. A retrospective analysis of this reporting method in comparison with films on a conventional viewing wall was performed with subjective and objective data. Radiologists and technicians were requested to evaluate the system on a subjective scale from 1 to 5 for seven topics (image analysis, inspection, time for reporting, handling, comparison with previous reports, quality of PC laser prints, training period). Moreover, personal interviews were conducted and protocols taken. Patient and image frequency, film costs, data transfer time and amount of system failures were evaluated as objective data. The most important results were: Two thirds of the film checking radiologists prefer the conventional viewing walls over the computer monitors. However, 70% of the residents prefer using the computer monitor for reporting. Seventy percent of the interviewed radiologists considered comparison with former examinations on film very difficult. Digitizing of former MRI examinations was not a convincing method; printing on a standard PC laser printer was considered to yield insufficient quality. The different acceptance between radiologists and residents seems to be related to different experiences. The reduction in film costs (48.6%) will improve further with complete PACS installation in the whole hospital. Data transfer rates are still poor; further improvement of network performance is necessary for convenient work. One whole MRI examination and report could be stored on CD-ROM for a cost of less than 2 euros. This could be a future means of cheap archiving and documentation suitable for viewing on any PC with DICOM III viewer. Images and reports could stay with the patients as in the past.