High volume instruments from different companies have been known to give different fiber strength for some varieties of cotton. A series of experiments were Conducted to identify some design features causing the problems. One experiment showed that by limiting hardware deflection during specimen loading the strength difference for some of the more troublesome cottons could be reduced significantly (from 1.79 to 0.4 gf/tex). A second experiment used mill processed samples to show that residual crimp in some fiber specimens can be a major source of error. A third experiment showed that by matching extension speeds the differences were reduced (from 1.72 to 1.45 gf/tex). An HVI experiment combining these effects gave a high correlation with universal fiber strength (R²=O.95) and demonstrated that varietal differences can be greatly reduced Dr eliminated when functional components are matched and mass measurements are improved.