Infiltrations with synthetic substances are effective strategies for repairing facial lipoatrophy. However, few data are available on long-term safety. We describe the safety of polyacrylamide hydrogel in 145 patients who received facial infiltrations with Aquamid from September 2002 to April 2004. Epidemiological, clinical (mainly complications), and psychological data (patient satisfaction) were collected. We also recorded all patients who presented with a local infection at any time after receiving an infiltration. Sixty-two percent of patients presented with severe facial lipoatrophy before infiltration. The cumulative volume of Aquamid injected was 5.5 ml (4-18) per patient. During a mean (SD) of 50.2 (4.3) months after infiltration, only one patient presented with a local infection. Small palpable, nonvisible nodules or indurations were the most frequent complications (19.3% and 6.2%, respectively). If we include the remaining patients from our center (n = 294) who also received Aquamid (although less than 4 years ago), a further three patients presented with a local infection (incidence of 0.9%). Most patients (88.9%) were "satisfied" or "very satisfied" with the results; patients with mild to moderate baseline facial lipoatrophy were more satisfied than those with severe lipoatrophy ("very satisfied": 92.7% versus 86.5%, respectively). Only 17.4% reported mild impairment of lipoatrophy and only 9.2% required new infiltrations; however, 76% would have preferred more infiltrations. The high patient satisfaction and the low number of severe complications after at least 4 years of facial infiltrations with Aquamid reflect the long-term safety of this product for the repair of facial lipoatrophy. However, prolonged follow-up of these patients is recommended to detect unexpected long-term adverse reactions.