The study of host plant adaptation in arthropods, and especially agricultural pests, presents great interest, as it relates to patterns of population isolation and differentiation, with effects ranging from gene flow restriction to speciation. Prompted by our earlier isoenzyme studies that revealed genetic differentiation of Tetranychus urticae (Koch) collected on citrus, compared to other host plants, we investigated crossing compatibility between T. urticae collected from citrus (lemon) trees (ELCI) and T. urticae collected from the weed Mercurialis annua (ELMA), in the same citrus grove. Crossing compatibility in haplodiploid species where unfertilised eggs develop into males, like T. urticae, is assessed based on both the number and the sex ratio of the offspring. When ELMA females were crossed with ELCI males, fecundity was not affected, however the sex ratio was significantly biased towards males (16.6 % females, compared to 66% in the control cross ELMAxELMA). In the reciprocal crosses (ELCIxELMA), fecundity was lower by 30% and the proportion of female offspring was reduced to 52% from 67 %, compared to the control. The fecundity of the F1 hybrid females was significantly reduced and the eggs they laid were less viable, compared to the non-hybrids, further reducing the reproductive potential of inter-strain crosses. Combined with previous data, these results suggest the existence of a citrus-associated T. urticae host race.KEY WORDS: host plant, reproductive incompatibility, Tetranychus urticae, two-spotted spider mite.