The length of the index finger relative to that of the ring finger, the 2D:4D ratio, has been taken to be a marker of the amount of testosterone (T) that was present in the foetal environment (Manning, Scutt, Wilson, & Lewis-Jones, 1998). It has also been suggested (Geschwind & Galaburda, 1987) that elevated levels of foetal T are associated with left-handedness and that adult levels of circulating T might relate to foetal levels (Jamison, Meier, & Campbell, 1993). We used multiple regression analyses to investigate whether there is any relationship between either left or right hand 2D:4D ratio and handedness. We also examined whether adult levels of salivary T (or cortisol, used as a control hormone) predict digit ratio and/or handedness. Although the 2D:4D ratio of neither the left nor the right hand was related to handedness, the difference between the digit ratios of the right and left hands, D(R-L), was a significant predictor of handedness and of the performance difference between the hands on a peg-moving task, supporting previous findings (Manning & Peters, 2009; Manning et al., 1998; Manning, Trivers, Thornhill, & Singh, 2000; Stoyanov, Marinov, & Pashalieva, 2009). Adult circulating T levels did not predict the digit ratio of the left or right hand; nor was there a significant relationship between concentrations of salivary T (or cortisol) and either hand preference or asymmetry in manual skill. We suggest that the association between D(R-L) and hand preference arises because D(R-L) is a correlate of sensitivity to T in the developing foetus.