There are now about fifty known radio pulsars in binary systems , including at least five in double neutron star binaries .
In some cases , the stellar masses can be directly determined from measurements of relativistic orbital effects .
In others , only an indirect or statistical estimate of the masses is possible .
We review the general problem of mass measurement in radio pulsar binaries , and critically discuss all current estimates of the masses of radio pulsars and their companions .
We find that significant constraints exist on the masses of twenty-one radio pulsars , and on five neutron star companions of radio pulsars .
All the measurements are consistent with a remarkably narrow underlying gaussian mass distribution , m = 1.35 \pm 0.04 M _ { \odot } .
There is no evidence that extensive mass accretion ( \Delta m \mathrel { \hbox to 0.0 pt { \lower 3.0 pt \hbox { $ \mathchar 536 $ } \hss } \raise 2 % .0 pt \hbox { $ \mathchar 318 $ } } 0.1 M _ { \odot } ) has occurred in these systems .
We also show that the observed inclinations of millisecond pulsar binaries are consistent with a random distribution , and thus find no evidence for either alignment or counteralignment of millisecond pulsar magnetic fields .