The aim of this study was to assess the up- and download performance of an image Web system (IWS) during simultaneous up- and download procedures. Following preparatory tests, the upload capacity (UC) with and without simultaneous download and the time-to-display (TTD) with and without simultaneous upload were determined for different image, server and compression types, and with up to 16 concurrent clients. The UC varied between 1.7-5.5 Gigabyte per hour (GB/h) and was slightly influenced by a simultaneous download; however, the TTDs were substantially prolonged during an upload and only below 5 s with up to four to six concurrent clients. The choice of image and server type had a strong impact on UC and TTD. Lossy primary compression proved slightly superior. An effective UC of approximately 4.0 GB/h or 96 Gigabyte per day (GB/day) can be achieved, which appears suitable even for large institutions. Because of its substantial influence on UC and TTD during simultaneous up- and download, the server hardware should be equipped with two processors and 1 GB RAM. Lossy primary compression may be used with slight performance benefits when full-resolution images are not required. The upload is a time-demanding process, and it is possible that during peak hours the waiting times are unacceptable for clinicians.