Over the years, many attempts have been made to develop fiber-to-yarn relationships for the sake of exploring the effects of fiber properties on yarn properties, and for the sake of predicting yarn characteristics. These attempts have met many challenges including: (1) a high degree of specificity making relationships strictly applicable to certain processes or particular yarns, (2) a great deal of instability in the coefficients associated with the relationships developed, and (3) a great deal of inaccuracy resulting from possible nonlinearities in some of the effects. As a result, no fiber-to-yarn models have been developed that can be considered as references for other applications. To make matters additionally complex, most of the relationships developed lacked the agreement between the known physical effects and the statistical effects. In this paper, we present a new approach in which universal fiber-to-yarn relationships can be established. This approach is unique in that it accounts for standard and known effects as well as actual data of fibers and yarn properties. Yarn and fiber parameters are expressed in the form of indices that can be converted into actual values depending on the specific application in question. The benefits of these relationships are numerous: (a) it will benefit cotton growers in establishing good realization of the technological values of cotton fibers, (b) it will assist greatly in exploring various effects of fiber properties on yarn characteristics for different spinning systems and end products, and (c) it will result in standardized fiber-to-yarn relationships.