We consider the spin and orbital evolution of low-mass binary radio pu
lsars which are not members of globular clusters. We show that their c
urrent spin periods can be understood in terms of their likely mass-tr
ansfer history. Systems with orbital periods greater than or similar t
o 50 days are the endpoints of low-mass X-ray binaries whose orbital e
volution is driven by the nuclear expansion of the secondary. For shor
ter periods, orbital angular momentum losses are significant, and the
mass-transferring progenitor undergoes a phase of Eddington or super-E
ddington mass transfer which spins up the neutron star to an Eddington
-limited equilibrium period. These systems subsequently enter a propel
ler phase and are probably undetectable as low-mass X-ray binaries, la
rgely resolving the inferred discrepancy between the current number of
binary pulsars and their likely progenitors. Spin-down of the neutron
stars in this propeller phase is evidently ineffective. The observed
relation between spin period, magnetic field, and orbital period is in
apparent agreement with simple ideas of the equilibrium spin period b
ut requires much lower fastness parameters than predicted by current t
heory.