We present numerical calculations of the photon-light-pseudoscalar-boson conversion in the recently discovered binary pulsar system J0737-3039. Light pseudoscalar bosons (LPBs) oscillate into photons in the presence of strong magnetic fields. In the context of this binary pulsar system, this phenomenon attenuates the light beam emitted by one of the pulsars, when the light ray goes through the magnetosphere of the companion pulsar. We show that such an effect is observable in the gamma-ray band since the binary pulsar is seen almost edge-on, depending on the value of the LPB mass and on the strenght of its two-photon coupling. Our results are surprising in that they show a very sharp and significant (up to 50%) transition probability in the gamma-ray (> tens of MeV) domain. The observations can be performed by the upcoming NASA GLAST mission.Light pseudoscalar bosons (LPBs) coupled to two photons have attracted considerable interest in the last few years and their implications for elementary-particle phenomenology, astrophysics and cosmology have been thoroughly investigated [1,2,3]. Specifically, the LPB mass m is assumed to be m < 1 eV and the LPBphoton-photon coupling is described by the interaction lagrangianwhere φ denotes the LPB field, F µν is the usual electromagnetic field strength (F µν is its dual), and M is the inverse coupling constant with the dimension of a mass. Natural Lorentz-Heaviside units with = c = 1 are employed throughout.Because the mass eigenstates of the LPB-photon system differ from the corresponding interaction eigenstates, interconversion takes place, a phenomenon quite similar in nature to flavour oscillations for massive neutrinos. However, since the off-diagonal term (1) involves "two photons", one of them actually corresponds to an external field. So, photon-LPB oscillations occur only in the presence of external electric or magnetic fields.A particular case of LPB is the standard axion, namely the pseudo-Goldstone boson associated with the PecceiQuinn U (1) global symmetry invented to solve the strong CP problem [4]. Although the exact value of the axion mass m is model-dependent, generally one finds m ≃ 0.7 (10 10 GeV/M ) eV. Besides providing a natural candidate for nonbaryonic dark matter [5], the standard axion can in principle be detected with high-precision laboratory experiments thanks to its two-photon coupling in eq. (1). In this connection, two strategies have been proposed. One is based on the resonant axion-photon conversion inside a tunable microwave cavity [6], while the other relies upon the induced ellipticity and change of polarization plane of an initially linearly-polarized laser beam [7]. In either case, a strong external magnetic field must be present. Generic LPBs with a two-photon coupling (1) can also be discovered in this way, provided of course that their mass m and inverse couplig constant M fall into suitable ranges which depend on the experimental arrangement.