The processes leading to the formation of early state societies remain one of the key topics of archaeological research. Few of these early states are as famous or evocative as that of ancient Egypt, a land of dramatic monuments and terrain, with mysterious and exotic religious practices and a distinctive and exotic iconography. But was Egypt the gift of the Nile, as the Greek historian Herodotus alleged? In this new book, Toby Wilkinson draws attention to a relatively neglected part of the Egyptian landscape: not the fertile river valley, but the deserts which fringe it to east and west. It is here in the deserts, he argues, that the origins of the Egyptian state are to be found. In recent millennia, the deserts have been hostile environments of rock and sand. Go back before 3000 bc, however, and a rather different picture emerges. This different picture is of a desert hinterland peopled by nomadic groups who spent part of their year in the Nile valley. It suggests a more mobile view of Egyptian Predynastic society than has usually been supposed. Desert and valley may have functioned together in a classic pattern of complementarity between contrasting environmental zones, with cattle herds perhaps moved from valley floor to desert in step with the cyclical pattern of the seasons. The specific ingredient which Wilkinson uses to link valley and desert during the fourth millennium bc is rock art. Egyptian rock art has not yet been properly recognized as a rich and important repertoire by specialists in the burgeoning field of rock art as whole. Surveys over more than a century, however, have revealed numerous groups of pecked and engraved images on the desert cliffs and boulders, and recent expeditions (including those by Wilkinson himself) are continually adding to the corpus. The Egyptian desert rock art is generally less well-known than the vivid rock paintings of the central Sahara (such as the famous Tassili frescoes), though it too conveys the image of a greener more habitable landscape. Wilkinson ties specific motifs found in the desert rock art to iconography from the Nile valley during the fourth millennium and later. Yet the linkages and chronologies remain controversial, along with the central hypothesis. Did the desiccation of the savannas lead to the formation of the Egypt, forcing the scattered pastoralist populations to withdraw to a cultivated Nile valley? Was Egypt the gift of the deserts, not the Nile? In this Review Feature the hypothesis is examined by specialists working in Egypt and Nubia, and the reliability of the supporting evidence is assessed.