It has been shown that some of the controversy over the structure of articular cartilage may be due to slight differences in the orientation of the sample that has been studied. As our decisive criterion we have used the simple physical fact that elongate proteins, such as collagen micelles, that can exhibit form-birefringence, had to show virtually straight extinction when viewed under crossed polars. The use of a variably adjustable microtome chuck facilitated small adjustments in the orientation of the cartilage to meet this criterion. Under these conditions, the collagen of the matrix has been shown to be aligned mainly perpendicularly to the surface which was bounded by a thin lamina in which the collagen showed birefringence at 90 degrees to that of the matrix. The conventionally described zonation of articular cartilage has been shown to be inadequate for that of the dog tibial plateau. The conventional Zone 2 has been shown to consist of two zones, Zones 2a and 2b, with different cell sizes, cell concentrations, and concentration of matrix components.