Long ago, Bohr, Pauli, and Mott argued that it is not, in principle, possib
le to measure the spin components of a free electron. One can try to use a
Stem-Gerlach type of device, but the finite size of the beam results in an
uncertainty of the splitting force that is comparable with the gradient for
ce. The result is that no definite spin measurement can be made. Recently t
here has been a revival of interest in this problem, and we will present ou
r own analysis and quantum-mechanical wave-packet calculations which sugges
t that a spin measurement is possible for a careful choice of initial condi
tions.