T he question of the rate of industrial growth is central to our understanding of the British industrial revolution. Thus, when Harley demonstrated that industrial growth in the period I770-i8I5 had been exaggerated by Hoffmann and by Deane and Cole, a powerful stimulus was given to interpretations which stressed both the slowness of overall growth and the elements of continuity with earlier experience.' Recent estimates by Crafts, Leybourne, and Mills (the CLM index) also show a low rate of industrial growth between I770 and i8I5 2 Together, the Harley and CLM estimates suggest the absence of any significant acceleration in the rate of growth of industrial production per head of population before i8I5. Earlier estimates by Crafts, in contrast, show a transition to faster industrial growth per head in the late eighteenth century.3 This article assesses these competing estimates of industrial growth and suggests that both the Harley and the CLM estimates are biased in ways which lead them to understate growth. The CLM index, in spite of its recent vintage, uses outdated and inadequate series to indicate growth in important industries. The estimates made by Crafts also have serious weaknesses, but there appears to be a smaller bias in his estimating procedure. Moreover, revisions which remove some of the easily remedied deficiencies do not substantially alter the pattern of longrun growth shown by Crafts's estimates. To this extent, his general picture of industrial growth is confirmed. This judgement, however, remains provisional because all the quantitative work in this area rests on a slender and fragile base. The article ends with a brief review of the principal difficulties which stand in the way of further attempts to reduce the extent of uncertainty about the rate of industrial growth.