The risk of cuckoo Cuculus canorus parasitism for great reed warbler Acrocephalus arundinaceus nests was evaluated in respect to nest and nest site structure in the reedbeds along canals in central Hungary, In total 90 nests were analysed, from which 31 remained unparasitised. 37 were single parasitised and 22 multiple parasitised. Multiple discriminant analysis, based on eight structural variables, separated well the unparasitised clutches from the cases of single and multiple parasitism, but the latter two categories also showed a weaker separation. Nests close to cuckoo vantage points are most vulnerable to parasitism. The variable “distance from nest to closest available cuckoo perch site” (tree or electric wire) played the most important role in the risk of parasitism. It showed highly significant differences among the groups: lowest distances were found in the case of multiple parasitism, and the highest distances were measured when nests remained unparasitised. Additionally, “nest visibility” also showed significant differences among the three groups: a more visible nest is more likely to be parasitised. Risk of parasitism affects on the host on two levels: 1) female cuckoos search for nest-building hosts from a perch site, but 2) when they are in the act of parasitism, they can find more visible nests in the reed. Besides the robust effects of the variables “distance to cuckoo perch site” and “nest visibility”, the variable “distance from nest to open water” may reduce, but “nest volume” increases the risk of multiple parasitism. Differences between cases of single and multiple parasitism are weak, mainly affected by chance. We explain it by the high abundance of the cuckoo, the parasitism rate was 65.6% (59/90), Cuckoos parasitised almost all of the available nests in the close vicinity of potential perch sites. There was no evidence that great reed warblers nested closer to each other when risk of parasitism was high.