Module averaging has been proposed by AMS as one way of improving the repeatability of HVI strength observations. AMS defines strength repeatability as "retesting of a bale's fiber strength should fall within 1.5 grams/tex of its original strength determination." For module averaging, AMS assumes that all the bales within a module have about the same true value and that the variation observed within modules is due mostly to sampling variability. Sampling variability is that variability one would obtain by repeated sampling of the same bale with the same instrument. The sampling variance of a module is reduced by a factor of 1/n or almost to zero. For example, if the sampling variance is 1.0 with 10 bales in a module, the variance is reduced to 0.1. However, if the true additional variability among bales is large, in our example greater than 1.08, then repeatability would be decreased. The within module variance of 13,000 Delta bales from about 1000 units usually was greater than the expected sample variance but not so large as to reduce repeatability. A retest of 418 bales from 31 modules and trailers showed individual bale repeatability as 6.7% and that for module average as 71.3%, indicating some improvement for module averaging. Module averaging has the disadvantage of requiring more work for the ginner, loss of quality control data, and may in some instances decrease repeatability. Advantages of module averaging are there may be a slight increase in repeatability, it limits bale selectivity, and the retesting of outliers is a good quality control practice.