Recent molecular analysis has proved the hypothesis that parts of the human sex chromosomes are homologous and can recombine in male meiosis. The biological consequences of this recombination have been investigated by considering the joint effect of neutral mutation and random genetic drift on a locus that is closely linked to the non-homologous segments of the sex chromosomes, but which recombines with them in an appreciable frequency. Our model predicts that, in the absence of selection, allelic differentiation between genes carried on the X and on the Y chromosomes will develop only if the recombination rate is of the same order of magnitude as the mutation rate or smaller. Similarly, a mutation favourable in males but disadvantageous in females will increase in frequency on the Y chromosomes, while remaining rare on the X chromosomes, only if the recombination rate is smaller than the fitness advantage of the mutation. The X and Y chromosomes are, thus, not expected to show any genetic differentiation for almost all of their homologous parts. Divergence will occur only for loci that very rarely recombine between the sex chromosomes.