Doubled haploids (DH) have become a standard tool in breeding and genetic studies of many crops and in most cases androgenesis is the only available route of their production. It has been recently observed that some populations of DH lines obtained via androgenesis contain high proportions of clones. This seriously reduces the efficiency of breeding and may jeopardize genetic studies. This study was designed to determine at which stage of androgenesis these clones are created, using samples set aside during routine production of DH lines in breeding of hexaploid triticale. The fate of each androgenic structure was carefully followed through the entire regeneration process, and all obtained plants were genotyped using DNA markers. Overall, 189 plants were regenerated forming 33 families, each originating from a single original androgenic structure (callus, polyembryos). In ca. 80 % of cases all members of a family were genetically identical. However, in about 20 % of cases the families of regenerants were genetically heterogeneous, showing that not all androgenic structures originate from single microspores. The evidence shown here demonstrates that retention of single plants from each original structure guarantees the production of only unique genotypes but it reduces the total output of plants. If maximum output is desired, multiple regenerants from single callus can be retained but must be genotyped using at least 10 polymorphic markers to identify clones.