A total number of 111 dogs were included in the present prospective study investigating the prevalence of Anaplasma phagocytophilum in dogs in Germany. Dogs were divided into two groups. Dogs of group 1 (n = 49) showed clinical and/or haematological signs seen in infections with A. phagocytophilum, whereas those of group 2 (n = 62) did not have any evidence of anaplasmosis. For each dog, an A. phagocytophilum 16S rRNA-nested polymerase chain reaction (PCR) of ethylenediaminetetraacetic acid (EDTA)-anticoagulated whole blood analysis, a microscopic evaluation of a buffy coat and a serum indirect fluorescent antibody test (IFAT) were performed. Forty-eight seroreactive dogs were identified altogether, which amounts to an overall point prevalence of 43.2%. There was no significant difference between the seroreactivity to A. phagocytophilum antigens among group 1 (44.9%) and 2 (41.9%) (P > 0.5). Seven dogs (6.3%) had positive PCR results. All of them were seroreactive. Six belonged to group 1. Morulae in neutrophilic granulocytes were found in two dogs of group 1 but in none of group 2. Both dogs were seroreactive. Very high antibody titres (> or =1:1024) were detected significantly more frequently in dogs with clinical signs attributable to infection with A. phagocytophilum (group 1) than in those without (group 2) (P < 0.001). There was no significant correlation of overall positives or antibody titres to age, breed, sex, or whether the dogs were family or working dogs. Dogs with high tick infestation were significantly more often seroreactive to A. phagocytophilum than those with no or low tick infestation (P = 0.007). In conclusion, there seems to be a high risk of infection with A. phagocytophilum in Germany. Results of this study suggest that severe illness solely caused by A. phagocytophilum may be possible although definitive evidence does not exist. Very high antibody titres (>1:1024) may be associated with clinical anaplasmosis.