The research work on Web-based long-distance data acquisition (DAQ) is valuable for application to tele-detection machine faults. With an expert system for machine fault detection, faults in a distantly located machine can be diagnosed through the internet. The distant user logs on to the expert system Web page, fills in the requirements, and starts-up the diagnose process. The system then connects to the DAQ server that is installed in the machine, samples data required for diagnoses through the internet, and sends back diagnose results. In such a long-distance system, Web-based DAQ plays an important role by automatic sampling and transferring of data through the internet. We have built an experimental data acquisition system using a National Instruments AT-MIO-16E-10 board running under Ch language environment. In this experimental example, the user can acquire data online. The principle of this experimental method is introduced in this paper. A detailed programming technique is described with an example.