According to the Geschwind-Behan-Galaburda (GBG) hypothesis, prenatal testosterone (T) causes a slowing in the development of the left brain with a consequent compensatory growth in the right brain, creating a reverse organisation of the cerebral lateralisation. That is, left- and right-handedness might be associated with high and low prenatal T levels, respectively. To test this hypothesis, the relations of T levels (umbilical cord blood) to grasp-reflex strengths were studied in human neonates. Handedness was assessed by measuring the grasp-reflex strengths from the right and left hands in 10 trials from each hand alternatively. There were two handedness groups: right-handers (R-L significantly greater than zero) and left- handers (significantly smaller than zero). Contrary to the GBG model, the mean free T concentration was found to be significantly higher in right-handers than left- handers for males and females. There was no significant difference in the total T levels between right-and left- handers. Free T concentrations positively correlated with R-L grasp-reflex strengths, i.e. right-handedness increased as T increased, and left- handedness increased as T decreased. Contrary to these positive correlations, T negatively correlated with the grasp-reflex strengths from the right and left hands. These results partly supported the GBG hypothesis for this spinal-motor-asymmetry model. Total T did not significantly correlate with grasp-reflex strengths. The results suggest that prenatal T may at least play a role in prenatal determination of spinal motor lateralisation, with a possible consequent upward regulation of cerebral lateralisation.