The extents of uptake of 0.5-15% of two disperse dyes on woven, knitted and hydroentangled nonwoven poly (ethylene terephthalate) fabrics were compared. The colour strength of the dyeings followed the order: nonwoven > knitted > woven. Differences in surface reflection were not responsible for the observed dye uptake behaviour, as the PET fibres used in the construction of the nonwoven, knitted and woven fabrics were of similar cross-sectional and longitudinal dimensions. The degree of PET fabric crystallinity could not explain the observed differences in colour strength and comparable melting temperatures were obtained for the three PET compositions. The rate of uptake of C.I. Direct Red 89 followed the order: nonwoven > knitted > woven cotton and also the colour strength of two direct dyes followed the order: nonwoven > knitted > woven and, in doing so, concurred with that of the two disperse dyes on the three types of PET fabric. The uptake of both disperse and direct dyes as a function of time was neither fibre- nor dye-related but was dependent upon some function of fabric construction. It was suggested that fundamental differences in fibre arrangement and porosity in woven, knitted and nonwoven fabrics resulted in differences in accessibility to dye molecules, according to which, whilst dye molecules may gain uniform access to all interior fibre surfaces in a hydroentangled nonwoven structure, only the outermost fibres within a yarn-based structure might be readily accessible. This was supported by the calculated porosity values for each undyed fabric type which followed the order: nonwoven > knitted > woven for both cotton and PET fabrics, which fitted well with the colour strength order: nonwoven > knitted > woven.