An account is given of an investigation of some factors influencing the extension of a woollen yarn during winding and of its subsequent relaxation. The work is reported in three main sections, (a) the effect of separate factors on extension, (b) the cumulative effect of successive winding tensions and (c) the retained extension of a stretched yarn after subsequent relaxation. The main results, for the yarn used, can be summarized as follows: Yarn removed from a bobbin on which it has been stored in a stretched condition retains some of the stretch which it had while on bobbin. As the time of storage on bobbin increases, up to about two weeks, the extension retained by the yarn also increases and thereafter remains fairly constant (up to four months). The extension retained when the yarn is unwound from a bobbin immediately after having been wound, is about three-quarters of the extension that might have been expected if the yarn had been stored for some weeks. Normal variations in regain do not cause very great differences in extension of yarns when wound under moderate tensions. When wound at the same tension, wet yarn stretches considerably more than the same yarn conditioned in the standard atmosphere. The pH of a previous treatment may have a small effect on the amount the yarn extends during winding. For the three winding speeds used, the extension retained decreases slightly with increase in winding speed. Yarns with different tension histories give different apparent extensions on rewinding at the same tension. The cumulative winding tension tests suggest a method of reducing the incidence of stripes and streaks by the use of a high winding tension at some stage of the processing. Relaxed yarn which has been stretched by different amounts and then soaked, untensioned, in water generally contracts to its initial relaxed length. In one case, a winding extension of 38½% was entirely recovered by subsequent relaxation.