BACKGROUND The treatment of severe bacterial infections of the ankle joint is difficult and complex. In the case of a chronic infection with destruction of the ankle joint, a tibiotalar arthrodesis with external fixation is the treatment of choice. In this study the results of ankle arthrodesis due to bacterial infection using the Ilizarov external fixator are presented. PATIENTS AND METHODS Between 2001 and 2004 37 patients (10 female, 27 male, mean age 58 years) were treated with a tibiotalar arthrodesis using the Ilizarov fixator. All patients had a confirmed infection in the course of their disease. Active infection was present in 20 patients at the time of the operation. Most secondary ankle arthritides (81 %) were caused post-traumatically after various internal fixation procedures. Previous ankle arthrodeses were tried in 14 cases (12 cases with internal fixation, two cases with external monolateral fixation). Patients were treated with a four-ring Ilizarov frame (in two cases with a five-ring frame) and stainless steel wires. All patients could be included at a mean follow-up of 46 (12-49) months. A modified AOFAS score was used for the functional outcome. RESULTS The operation took 141 minutes at an average ranging from 90 to 252 minutes. The inpatient treatment lasted between 10 and 63 days (mean 26 days). The time spent in the fixator was 116.7 (69-245) days. All patients were mobilised under full weight bearing with the external fixator. Surgical revision was necessary in 13 patients: four patients needed wound revisions due to ongoing infection, six patients needed wire exchange due to deep infection in three cases and wire breakage in three cases, one patient needed additional wires because of an initially instable frame configuration and two patients needed secondary skin grafting. Bony consolidation was achieved in 32 patients (86.5 %). With a re-arthrodesis performed in four patients using the Ilizarov fixator, the overall fusion rate was 94.6 %. Infection was persistent in two cases with one solid ankle fusion and one ankle pseudarthrosis. At the time of follow-up 35 patients were able to walk under full weight loading with orthopaedic shoe modifications, four patients needed support of a cane and three patients wore an ankle-foot orthesis. The two patients with persistent pseudarthrosis were mobilised in a lower-leg orthesis after declining another surgical revision. The positioning of the hindfoot showed in seven cases an equinus of 10°, in one case a varus of 10° and in two cases a valgus positioning of 10°. A plantigrade foot positioning or with minimal degrees of deviation could be achieved in all other cases. The modified AOFAS score at the time of the follow-up examination ranged from 19 to 86 with an average score of 67.9 points. All patients except three were satisfied or rather satisfied with the treatment procedure and its results. CONCLUSION The Ilizarov external fixator is a safe method for ankle fusion in cases of infection. The advances are a possible application at acute infection and immediate mobilisation at full weight bearing. However, it remains a time-consuming and stressful procedure for the patient.