This paper presents the energy balance for two specific days during the summer of 1991 at the camp of the Vrije Universiteit Amsterdam (VU-Camp) on the Greenland ice sheet. On the basis of this energy balance the amount of melt for both days is estimated. This amount is compared with the melt as measured at the position of two snow stakes. Although numerous difficulties are associated with the interpretation of stake measurements, both methods are seen to agree within experimental error.