Photometric standard observations are routinely obtained by ESO, and nightly zero points are computed mainly for the purpose of monitoring the instrument performance. The accuracy of these zero points is about 10%.

Recently, we have started a program to investigate, if and how percentlevel absolute photometric accuracy with FORS can be achieved. The main results of this project are presented in this paper. We first discuss the quality of the flatfields and how it can be improved. We then use data with improved flatfielding to investigate the usefulness of Stetson standard fields for FORS calibration and the accuracy which can be achieved.

The main findings of the FORS Absolute Photometry Project program are as follows. There are significant differences between the sky flats and the true photometric response of the instrument which partially depend on the rotator angle. A second order correction to the sky flat significantly improves the relative photometry within the field. Percent level photometric accuracy can be achieved with FORS1. To achieve this accuracy, observers need to invest some of the assigned science time for imaging of photometric standard fields in addition to the routine nightly photometric calibration.