Amplified fragment length polymorphisms (AFLP) are dominant markers frequently used to build linkage maps where heterozygosity could be inferred by a backcross breeding strategy. In the present study, we describe the utilization of an unmanipulated great reed warbler, Acrocephalus arundinaceus pedigree to infer heterozygous genotypes of AFLP markers in order to map these markers to a partial linkage map previously based on microsatellites. In total, 50 of the 83 autosomal AFLPs (60%) and 4 of 5 Z-linked AFLPs (80%) were mapped. For each marker, on average, 88% of the expected number of heterozygote parents was detected. The likelihood of map assignment was to a large extent due to the number and density of microsatellite markers already in the map. The 'parsimonious linkage map', that is the map based on the most parsimonious location of all significantly linked markers, consisted of 21 autosomal linkage groups with 2 to 15 markers and had a total map size of 552 cM in males and 858 cM in females. The Z-chromosome linkage group with 12 markers had a size of 155 cM. The autosomal 'framework linkage map', that is the map based only on markers with an unambiguous position, had a total size of 237 cM in males and 440 cM in females, respectively. The inclusion of AFLPs enlarged the previous map substantially (e.g. the autosomal parsimonious linkage map became 441 cM and 621 cM larger for male and female recombination, respectively). The probability that an AFLP became mapped increased with increasing level of heterozygosity, whereas the probability of mapping into a framework position increased with both heterozygosity and number of genotyped individuals. Our results suggest that AFLP provides a fast and inexpensive means of enlarging genetic maps already composed of markers with high polymorphism, also in wild populations with unmanipulated pedigrees.