The study was undertaken to show whether developmental homeostasis is linked with heterozygosity in organisms with haplo-diploid systems. Developmental homeostasis is defined in its broadest sense as the ability of individuals to resist disturbances during development by reacting in a buffered way (Lerner, 1954). Disturbances during the development of an individual will be reflected in morphological variation of the adult; homeostatic abilities will lead to a reduction of such variation. This study deals with developmental homeostasis in honeybees as expressed in their adult body symmetry. It has been found that heterozygous individuals are in general superior in their homeostatic ability to homozygous individuals. Lewontin (1956) found for example that Drosophila from lines which were made homozygous for different second chromosomes showed less correlation in abdominal bristle number between the 4th and 5th sternites than flies from heterozygous lines. The fact that the anteriorposterior gradient in bristle number was better expressed in the heterozygous individuals was interpreted as being due to their superior buffering ability during their development. The interactions of genes and alleles in the heterozygous genomes are thought to improve this buffering ability, i.e. the homeostatic development of the individuals. Lewontin found that heterozygous flies had higher average adaptive values over a range of environments than homozygous flies. He concluded that homozygous flies might occasionally be better in one special environment but that they are inferior to heterozygous individuals over an array of environments. The relative importance of special coadapted systems versus general heterosis is not yet known. Former research on homeostasis has been done on diploid organisms only, mainly on Drosophila (Dobzhansky and Levene, 1955; Lewontin, 1956, 1958; Lerner, 1954; Thoday, 1953). This study deals with homeostasis in a haplo-diploid system to find out if the improvement of homeostasis due to heterozygosity is also present in a species where part of the population has only half the diploid chromosome set. The honeybee was chosen because it has such a system: all males (the drones) are hemizygous, developing from unfertilized eggs and all females (the workers and queens) are heterozygous, developing from fertilized eggs. Since it is possible to inseminate the queens artificially, one can get workers with different degrees of homozygosity by inbreeding queens. It should be mentioned here that it is impossible to get completely homozygous workers because of the sex-determining mechanism in honeybees: there is no X or Y chromosome but a series of multiple alleles at one locus that determine the sex -if it is heterozygous, workers or queens will develop; if it is homozygous or hemizygous, drones will develop (Mackensen, 1951). Diploid drones do not exist in normal colonies, because they are removed by the workers in early stages of their development. However, in the laboratory it has been shown that drones will develop from diploid eggs in which the sex-locus is homozygous (Woyke, 1965). 1 This paper is part of a doctoral dissertation at the University of Munich.