This preliminary data suggests that knitting may benefit inpatients with eating disorders by reducing their anxious preoccupations about eating, weight and shape control. The specificity of this effect is yet to be determined. This preliminary outcome requires further controlled study in AN subjects. From a clinical perspective, knitting is inexpensive, easily learned, can continue during social interaction, and can provide a sense of accomplishment. The theoretical and empirical rationale for this observation, and implications for deriving alternative strategies to augment treatment in AN, are discussed.