This article reexamines theories previously advanced to explain Lancashire's slow adoption of ring spinning. New cost estimates show that although additional transport costs and technical complementarities reduced ring adoption rates, these supply-side constraints were not dominant. Instead what mattered most were demand-side factors. Lancashire produced far more fine yarns than other countries, and this yarn was better spun on mules. Furthermore, Lancashire had a sizeable export trade in yarn, a market again more suited to mule spinning. Low ring adoption rates were thus a rational response to demand patterns dominated by high-quality goods.