Lipid membranes and fiber networks in biological systems perform important mechanical functions at the cellular and tissue levels. In this thesis I delve into two detailed problems -- thermal fluctuation of membranes and non-linear compression response of fiber networks. Typically, membrane fluctuations are analysed by decomposing into normal modes or by molecular simulations. In the first part of my thesis, I propose a new semi-analytic method to calculate the partition function of a membrane. The membrane is viewed as a fluctuating von Karman plate and discretized into triangular elements. Its energy is expressed as a function of nodal displacements, and then the partition function and co-variance matrix are computed using Gaussian integrals. I recover well-known results for the dependence of the projected area of a lipid bilayer membrane on the applied tension, and recent simulation results on the ependence of membrane free energy on geometry, spontaneous curvature and tension. As new applications I use this technique to study a membrane with heterogeneity and different boundary conditions. I also use this technique to study solid membranes by taking account of the non-linear coupling of in-plane strains with out-of-plane deflections using a penalty energy, and apply it to graphene, an ultra-thin two-dimensional solid. The scaling of graphene fluctuations with membrane size is recovered. I am able to capture the dependence of the thermal expansion coefficient of graphene on temperature. Next, I study curvature mediated interactions between inclusions in membranes. I assume the inclusions to be rigid, and show that the elastic and entropic forces between them can compete to yield a local maximum in the free energy if the membrane bending modulus is small. If the spacing between the inclusions is less than this local maximum then the attractive entropic forces dominate and the separation between the inclusions will be determined by short range interactions; if the spacing is more than the local maximum then the elastic repulsive forces dominate and the inclusions will move further apart. This technique can be extended to account for entropic effects in other methods which rely on quadratic energies to study the interactions of inclusions in membranes. In the second part of this thesis I study the compression response of two fiber network materials -- blood clots and carbon nanotube forests. The stress-strain curve of both materials reveals four characteristic regions, for compression-decompression: 1) linear elastic region; 2) upper plateau or softening region; 3) non-linear elastic region or re-stretching of the network; 4) lower plateau in which dissociation of some newly made connections occurs. This response is described by a phase transition based continuum model. The model is inspired by the observation of one or more moving interfaces across which densified and rarefied phases of fibers co-exist. I use a quasi-static version of the Abeyaratne-Knowles theory of phase transitions for continua with a stick-slip type kinetic law and a nucleation criterion based on the critical stress for buckling to describe the formation and motion of these interfaces in uniaxial compression experiments. Our models could aid the design of biomaterials and carbon nanotube forests to have desired mechanical properties and guide further understanding of their behavior under large deformations.