In the mouse, gametes with gross chromosome duplications and deficiencies can complement each other to give viable zygotes (with some notable exceptions involving particular chromosomes). These complementation-type offspring can be recognised in intercrosses between translocation heterozygotes in which one parent is homozygous for a recessive genetic marker not carried by the other. This system has beeN used by Lyon and colleagues (1976) to study non-disjunction in heterozygotes for tobacco mouse and laboratory-derived Robertsonian translocations. Although non-disjunction is frequent in the former group, still higher frequencies are needed for a workable test system in which wild type mice are treated and mated to a tester stock generating many aneuploid gametes. Possible approaches include (1) use of semidominant markers, (2) marking both arms, (3) combining two or three independent Robertsonians in the tester stock, (4) use of compounds of Robertsonians wih monobrachial homology, since these give very high frequencies of non-disjunction, (5) generation of a compound of three Robertsonians with tribrachial homology, which should produce aneuploid gametes only. This last seems the most promising approach, if the compound proves fertile, and would be analogous to the isochromosome system of Drosophila.