Background. Multifocal electroretinograms (mfERG) were recorded using a confocal scanning laser ophthalmoscope (cSLO) and compared to the results from conventional monitor stimulation. Methods. Single and repeated measurements were recorded from 23 normal subjects using the cSLO (Heidelberg Retina Angiograph, Heidelberg Engineering, Heidelberg) as well as a conventional monitor as stimulation devices. Laser power output was modified by various optical filters. The reliability of the method and agreement with the conventional monitor stimulation were determined. Results. CSLO recordings showed a high degree of variability. Reduction of laser power output improved the retinal response topography and characteristically modified response variations with each filter. Differences in amplitude size between cSLO and monitor recordings decreased with increasing amplitude levels. The results of repeated measurements showed considerable variation. Conclusion. It is possible to use a cSLO as a stimulator for mfERG recordings. However, a relatively high degree of variability represents a significant limitation of this method. Appropriate reduction of laser power decreases variations and serves to obtain photopic response topographies.