To process large amounts of data, in some fields of science hundreds or thousands of single jobs are submitted into a Grid. Monitoring the enormous numbers of jobs and their resource usage in such environments (like the LCG/gLite middleware) effectively becomes an important issue for the users. Current tools in LCG / gLite provide only limited value to the user as they are often simple command line applications only. Keeping an eye on large number of jobs can thus become quite painful. In this paper, the user-centric monitoring system AMon is presented that has been developed within the HEPCG project. It provides Grid users with useful and graphical information that helps her/him to understand the status of the jobs and their usage of resources. The latter enables the user to better recover from job failures.