Evidence is presented to support the recent suggestion that human neural tube closure is similar to that observed in mice, and comprises several regionally distinct closure sites rather than being a simple zipping up process. Seven subjects, each with more than one neural tube defect (NTD), are described. Comparative studies of the location of the lesions in relation to the closed parts of the neural tube imply that such NTD could only have come about if there were intermittent closure of the neural tube.