AbstractIt is shown that while the Bowditch adjustment is essentially a shearing process which may produce large corrections to the first and last angles of a traverse, the loss in precision, when compared with a least-squares solution, is not serious. Furthermore when there are large differences between the Bowditch and least-squares corrections to the coordinates of stations of L-shaped and square traverses, the least-squares solution is sensitive to the ratio of the variance of distances to that of angle measurement, used in the adjustment.