Despite the ubiquity of coinfection, we know little of the effects of intra-specific genetic variability on coinfection by distinct parasite species. Here we test the hypothesis that parasite multiplication depends on the combination of parasite genotypes that coinfect the host (that is Genotype.parasite × Genotype.parasite interaction). To that aim, we infected tomato leaves with the ecto-parasitic mites Tetranychus urticae and Tetranychus evansi. We tested all possible combinations between four T. urticae and two T. evansi populations sampled on different hosts or localities. There was no universal (that is genotype-independent) effect of coinfection on mite multiplication; in many cases the two species had no effect on each other. However, several combinations of T. evansi and T. urticae populations led to elevated T. evansi numbers. Similarly, T. urticae reproduction largely depended on the interaction between T. urticae and T. evansi populations. This evidence for genotype-by-genotype interaction between coinfecting parasites indicates that the effect of coinfection on parasite epidemiology and evolution may vary in space according to the genetic composition of local parasite populations; it further suggests the possibility of coevolution between parasites species that share the same hosts.