In many respects, this paper may be regarded as a logical sequel to an earlier one presented to the Institute nearly seven years ago. In that paper the assessment of industrial ordinary shares was discussed in considerable detail and methods were suggested which could be used both to make projections of future dividends and earnings for different companies and to utilize these estimates in making investment decisions. Much of the discussion on the earlier paper centred on the latter problem and the present paper is concerned entirely with this aspect of share assessment. The making of future projections will not be considered again; suffice it to say that our methods have remained largely unchanged. The principle on which the technique that we have now developed rests is closely linked with the previous concept of a ‘capitalization rate’ but in passing we would mention that we have made a brief experiment with what we term ‘growth yields’, i.e. the compound interest return to be expected from a particular share at a given price on the basis of the forecast future dividends. Our results utilizing this method were in fact significantly better than those obtained using the more familiar concept of relative yields (see §8) but, as the figures presented later in this paper show, they were not as good as those obtained using the method which forms the basis of this paper.