The T-matrix method is used numerically to calculate the light scattering patterns of monodisperse, randomly oriented, non-spherical biological aerosol particles based on prolate spheroidal and spherical models in order to obtain size and shape sensitive information. The shape dependence of the phase function for the biological aerosol particles is discussed. The angular distributions of light scattered by Staphylococcus, Yersinia pestis, Francisella tularensis, Shigella dysenteriae and Burkholderia pseudomallei are presented and the size and shape dependence of the scattering matrix elements are illustrated in detail. At the forward- and side-scattering angles, the phase function has little dependence on particle shape for particles with equal extinction cross-section. It indicates that spherical model can be used to model non-spherical particles at these regions. The ratio of integrated backward- scattering at 170°～180°of F11(θ) to integrated forward-scattering at 0°～10° increases with the increases of the aspect ratio for non-spherical particles, which can be used to characterize particle shape. For aerosols of different sizes and shapes, the forward-scattering steepness ［F11(0°)-F11(5°)］/5 of F11(θ) increases with surface-equivalent radius, which provides theoretical basis for particle size detection. The ratio of integrated scattering at 170°～180°to integrated scattering at 0°～10°of F11(θ) as well as integrated scattering at 30°～90° of F22(θ)/F11(θ) increases with the aspect ratio, which can be used to determine the shape of particles. The study results in this paper provide a theoretical foundation to the design of particle size and shape analysis apparatus and the fast and effective detection of harmful biological micro-organisms in the air.