Data on the development of oscillatory ripple marks under waves are utilized as a further check on the equations that have been proposed for the threshold of sediment motion. If the threshold curve is correct, the data on ripple occurrence should lie in a stress field above the curve for threshold. Analysis of the data shows this mainly to be true but suggests that the curve for the threshold should be lowered slightly to smaller stress values as a number of ripple occurrences would otherwise fall below the threshold curve. As the stress of the wave orbital motion increases, the ripple heights progressively decrease and ultimately disappear at a critical stress value. Data on this disappearance and plane-bed sheet sand transport development are examined and compared with the data on the high-stress presence of ripples. There is good agreement of the two data sets and also confirmation of the theoretical criterion of Bagnold for the disappearance of ripples.