This paper gives a short introduction in the history of the determinat
ion of atomic masses. Their differences can often be found with high p
recision from measurements of nuclear and reaction energies. But here,
emphasis is given to features determining the precision with which th
e masses themselves can be determined with mass spectrometers.