Neurulation refers to the folding process in vertebrate embryos, which includes the transformation of the neural plate into the neural tube. The embryo at this stage is termed the neurula.

The process begins when the notochord induces the formation of the central nervous system (CNS) by signaling the ectoderm germ layer above it to form the thick and flat neural plate. The neural plate folds in upon itself to form the neural tube, which will later differentiate into the spinal cord and the brain, eventually forming the central nervous system. Computer simulations found that cell wedging and differential proliferation are sufficient for mammalian neurulation.

Different portions of the neural tube form by two different processes, called primary and secondary neurulation, in different species.

The concept of induction originated in work by Pandor in 1817. The first experiments proving induction were attributed by Viktor Hamburger to independent discoveries of both Hans Spemann of Germany in 1901 and Warren Lewis of the USA in 1904. It was Hans Spemann who first popularized the term "primary neural induction" in reference to the first differentiation of ectoderm into neural tissue during neurulation. It was called "primary" because it was thought to be the first induction event in embryogenesis. The Nobel prize-winning experiment was done by his student Hilda Mangold. Ectoderm from the region of the dorsal lip of the blastopore of a developing salamander embryo was transplanted into another embryo and this "organizer" tissue "induced" the formation of a full secondary axis changing surrounding tissue in the original embryo from ectodermal to neural tissue. The tissue from the donor embryo was therefore referred to as the inducer because it induced the change. While the organizer is the dorsal lip of the blastopore, this is not one set of cells, but rather is a constantly changing group of cells that migrate over the dorsal lip of the blastopore by forming apically constricted bottle cells. At any given time during gastrulation there will be different cells that make up the organizer.

Subsequent work on inducers by scientists over the 20th century demonstrated that not only could the dorsal lip of the blastopore act as an inducer but so could a huge number of other seemingly unrelated items. This began when boiled ectoderm was found to still be able to induce by Johannes Holtfreter. Items as diverse as low pH, cyclic AMP, even floor dust could act as inducers leading to considerable consternation. Even tissue which could not induce when living could induce when boiled. Other items such as lard, wax, banana peels and coagulated frog's blood did not induce. The hunt for a chemically based inducer molecule was taken up by developmental molecular biologists and a vast literature of items shown to have inducer abilities continued to grow.

More recently, the inducer molecule has been attributed to genes, and in 1995, there was a call for all the genes involved in primary neural induction and all their interactions to be catalogued, in an effort to determine "the molecular nature of Spemann's organizer". Several other proteins and growth factors have also been invoked as inducers, including soluble growth factors such as bone morphogenetic protein and a requirement for "inhibitory signals" such as noggin and follistatin.

Even before the term induction was popularized, several authors, beginning with Hans Driesch in 1894, suggested that primary neural induction might be mechanical in nature. A mechanochemical-based model for primary neural induction was proposed in 1985 by G.W. Brodland and R. Gordon. An actual physical wave of contraction has been shown to originate from the precise location of the Spemann organizer which then traverses the presumptive neural epithelium and a full working model of how primary neural inductions was proposed in 2006. There has long been a general reluctance in the field to consider the possibility that primary neural induction might be initiated by mechanical effects. A full explanation for primary neural induction remains yet to be found.

As neurulation proceeds after induction, the cells of the neural plate become high-columnar and can be identified through microscopy as different from the surrounding presumptive epithelial ectoderm (epiblastic endoderm in amniotes). The cells move laterally and away from the central axis and change into a truncated pyramid shape. This pyramid shape is achieved through tubulin and actin in the apical portion of the cell which constricts as they move. The variation in cell shapes is partially determined by the location of the nucleus within the cell, causing bulging in areas of the cells forcing the height and shape of the cell to change. This process is known as apical constriction. The result is a flattening of the differentiating neural plate which is particularly obvious in salamanders when the previously round gastrula becomes a rounded ball with a flat top. See Neural plate.