This paper examines sperm competition games in which both size and number of gametes in an ejaculate can be varied strategically, but under the control of the haploid sperm. It is a companion to the previous paper (Parker (Proc. R. Soc. Lond. B 253, 245 (1993))) for the case where the strategic control is by the diploid parent. Under haploid control, the optimal balance between size and number of sperm in an ejaculate becomes quite different from that predicted for diploid control; there is a conflict between parental and gametic interests over sperm size and number. This occurs because both intra-ejaculate and inter-ejaculate sperm competition are present with haploid control, whereas intra-ejaculate competition is absent under diploid control. The magnitude of this conflict is affected by the risk of inter-ejaculate sperm competition, and by the fact that gametic interests depend on the nature of available mutations. If ejaculate expenditure trades off against numbers of matings achieved, with haploid control there is escalation towards maximal expenditure on the ejaculate. If the ejaculate expenditure is fixed by the diploid parent, but there is a size-number trade-off under haploid control, solutions depend on the mode of action of the mutation affecting sperm size or number. If mutant-bearing sperm deviate entirely at the expense of non-mutant sperm, any increase in size or number will spread. If mutant sperm deviate entirely at their own expense, the evolutionarily stable strategy (ESS) is the same as for diploid control; there is no conflict. If mutant and non-mutant sperm compete for limited resources, the ESS depends on the type of mutation: (i) mutations affecting sperm size give very large sperm, decreasing as inter-ejaculate sperm competition increases; (ii) mutations affecting sperm number give very small sperm, increasing as inter-ejaculate sperm competition increases. Conflict between diploid and haploid expression is reduced by inter-ejaculate sperm competition, but is not lost, even under maximum competition. We suggest that size and other morphological variation of sperm may sometimes reflect the result of conflict between different mutational types under haploid expression, and their conflict with diploid expression.