Around 1637, the French mathematician Pierre de Fermat wrote that he had found a way to prove a seemingly simple statement: while many square numbers can be broken down into the sum of two other squares - e.g. 25 (five squared) equals 9 (three squared) plus 16 (four squared) - the same can never be done for cubes or any higher powers. Fermat's formula did not survive, however, and some of the world's greatest mathematical brains have since worked long and hard to work out how he might have done it. In 1993, after a seven-year struggle, Andrew Wiles presented to an astonished conference in Cambridge a 200-page proof using algebra, analysis, geometry and topology as well as number theory. A fatal flaw was discovered and then brilliantly overcome. Here, weaving together history and science, Amir D. Aczel offers a thrilling, step-by-step account of the search for the mathematicians' Holy Grail.