The cerebrum of mammals spans a vast range of sizes and yet has a very regular structure. The amount of folding of the cortical surface and the proportion of white matter gradually increase with size, but the underlying mechanisms remain elusive. Here, two laws are derived to fully explain these cerebral scaling relations. The two general laws are derived from the notion that total processing power of the cortex is determined by the total cortical surface (i.e., the number of neurons), whereas the most efficient over-all flow of information is governed by the size of local networks (cortical columns). Since information is transferred by axonal connections which have a definite volume, a trade-off can be formulated from theoretical considerations between local, inter-gyral information transfer and long-range information transfer. It can be shown that this trade-off is governed by a single parameter describing the size of local networks, t local . Despite having just one free parameter, the first law fits the mammalian cerebrum better than any existing function, both across species and within humans. According to the second law, the scaling of white matter volume is also determined by the information principles. It follows that large cerebrums have much local processing and little global information flow. Moreover, paradoxically, a further increase in long-range connections would decrease the efficiency of information flow. These theoretical scaling principles help to compare the cerebrums across mammals regardless their size.