Converted head waves are observed sufficiently often in crustal and other refraction surveys to raise interesting questions as to their possibilities in interpretation. Some simple cases are examined in which sequences of P, S, and converted head waves are assumed to be observed over a two- and a three-layer medium. Investigation is made of the theoretical requirements in terms of numbers of shot and recording points, and the number of members in the sequence, to allow calculation of layer velocities and thicknesses when there is low to moderate structure on the interfaces.It is found that as few as a single shot and a single recording point, and two, three, or four members of a sequence are sufficient to allow determination of some or all of the P and S velocities in the layers, and the depths to the layers below shot and recording point. How far these possibilities can be realized in practice is dependent on the development of techniques to produce and observe converted waves consistently in refraction surveys. To what extent this can be done is at present little known, and is a matter for future research. If converted waves can be used regularly in refraction work, some operational advantages might result as follows. Because of the decrease in the number of shot and recording points necessary to derive velocities and depths when sequences of head waves are observed, it is possible either to gather usable data with minimum effort, or, if a more elaborate array of points is used, to achieve greater detail regarding the lateral variation in velocity and depth over what would have been obtained from a single wave type.