Lattice materials (LM) are a novel concept stemming from the combination of crystallography and structural optimisation algorithms. Their practical applications have become real with the advent of versatile additive layer manufacturing (ALM) techniques and the development of dedicated CAD/CAE tools. This work critically reviews one of the major claims concerning LMs, namely their excellent stiffness-to-weight performance. First, a brief literature review of spatially uniform LMs is presented, focusing on specific strength of standard engineering materials as compared with novel structures. An original modelling and optimisation is carried out on a flat panel subject to combined shear and bending load. The calculated generalised specific stiffness is compared against reference values obtained for a uniform panel and the panel subjected to topological optimisation. The monomaterial, a spatially repetitive solution turns out to be poorly suited for stiff, lightweight designs, because of suboptimal material distribution. Spatially non-uniform and locally size-optimised structures perform better. However, its advantage over manufacturable, topologically-optimised conventional designs can at best be marginal (< 10%). Cubic-cell lattices cannot replace conventional bulk materials in the typical engineering use. The multi-cell-type and multi-material lattice structures, albeit beyond the scope of this article, are more promising from the point of view of mechanical properties. The possibility of approaching the linear scaling reported in the recent litterature can make them an attractive option in ultra-low weight designs.