This paper presents the implementation of two optical methods for characterizing the state of fabric surfaces: a multidirectional roughness meter and a hairiness meter. A textile fabric has a complicated texture in that the relevant components (fibers) are not very small (usually at least 10 ?m in diameter), but in the fabric are partly ordered (in the yarn) and partly disordered (in the superficial hairiness). Hairiness is important in giving a textile fabric a characteristic surface. It makes the effective surface three-dimensional with many hidden areas. Thus, characterizing the state of a fabric surface requires special devices. On the one hand, the presented roughness meter measures essentially the fabric-reflected rays in all surface directions, because the fabric is rotated in its plane during the measurement. A Fourier temporal analysis of the reflected beam scanned across the fabric surface allows the fabric structure periods to be determined, because these periods yield peaks in the frequency spectrum. Two examples are given that show that the peak heights characterize the state of the fabric surface. On the other hand, the hairiness meter measures the hairiness of the fabric surface by an edge extraction method, which separates hairiness and structural information. The two devices are complementary and allow the processes that modify the fabric surface to be understood and controlled.