Synthetic polarization and flux spectra are presented for aspherical, electron scattering-dominated photospheres of Type II supernovae (SN II) in general and the specific case of SN 1993J. Monte Carlo calculations are based on the following assumptions: a) ellipsoidal envelopes with power law density profiles; b) occupation numbers given by local thermodynamical equilibrium (LTE); c) pure electron scattering for continuum opacities; d) lines treated in a Sobolev approximation with an assumed constant thermalization fraction; e) line transitions result in depolarization; and f) the temperature structure is given by a gray extended atmosphere.The observed luminosity of a Type II supernova depends on the unknown inclination angle i. Spectral analysis alone will fail to detect even strong deviations from spherical symmetry. Line scattering depolarizes incident polarized light, but the residual intrinsic polarization does not completely vanish because of electron scattering effects that depend on the electron density distribution. By combining results on the polarization and velocity structure of the emission lines, and the degree of polarization in the continuum, we place strong constraints on the degree of asphericity, the inclination of the system and the electron density distribution. In addition, since the problem is over constrained, we can independently test for the contribution to the polarization caused by aligned interstellar grains between the SN and earth.These modelling techniques have been applied to SN1993J. Both the flux spectra and the percentage of polarization as a function of wavelength can be reproduced by an aspherical model with an axis ratio of 0.6, radial electron densities proportional to r −5 and an effective temperature of 4800 K. In an oblate model, SN 1993J is seen almost equator on. The line forming region is still within the hydrogen rich part of the envelope 3 weeks after the explosion, although the continuum may form in deeper layers. Although the residual polarization across Hα is not zero, the interstellar component derived herein is consistent with that deduced by Trammell et al. (1993). The power of this technique for investigating the structure of other SN II, as a method for independently deriving the interstellar polarization, and the implications of these results on the use of SN II to determine distances through the Baade-Wesselink method are discussed.-2 -