BACKGROUND: The skin disorder Post Kala-Azar Dermal Leishmaniasis (PKDL) occurs in up to 10% of patients treated for visceral leishmaniasis (VL) in India. The pathogenesis of PKDL is not yet fully understood. Cases have been reported in India following therapy with most available treatments, but rarely in those treated with liposomal amphotericin B (Ambisome). Between July 2007 and August 2012 with the support of the Rajendra Memorial Research Institute (RMRI), Médecins Sans Frontières (MSF) supported a VL treatment programme in Bihar, India-an area highly endemic for Leishmania donovani-in which 8749 patients received 20 mg/kg intravenous Ambisome as first-line treatment. This study describes the characteristics of patients who returned to the MSF supported treatment programme with PKDL. METHODS AND PRINCIPAL FINDINGS: Over a 5-year period, Ambisome was administered to 8749 patients with laboratory-confirmed VL (clinical signs, rK39 positive, with/without parasite confirmation) in four intravenous doses of 5 mg/kg to a total of 20 mg/kg, with a high initial-cure rate (99.3%) and low default rate (0.3%). All patients received health education highlighting the possibility and symptoms of developing PKDL, and advice to return to the MSF programme if these symptoms developed. This is an observational retrospective cohort study of the programme outcomes. Of the 8311 patients completing treatment for their first episode of VL, 24 (0.3%) returned passively to the programme complaining of symptoms subsequently confirmed as PKDL, diagnosed from clinical history, appearance consistent with PKDL, and slit-skin smear examination. Of the 24 patients, 89% had macular lesions, with a median time (interquartile range) to development of 1.2 (0.8-2.2) years following treatment. Comparison of the demographic and clinical characteristics of the VL patients treated with Ambisome who later developed PKDL, with those of the remaining cohort did not identify any significant risk factors for PKDL. However, the time to developing PKDL was significantly shorter with Ambisome than in a subset of patients presenting to the programme with PKDL following previous sodium stibogluconate treatment for VL. CONCLUSIONS: In this large cohort of patients with VL in Bihar who were treated with 20 mg/kg Ambisome, PKDL following treatment appears to be infrequent with no predictive risk factors. The shorter median time to developing symptoms of PKDL compared with that after conventional VL treatments should be taken into account when counseling patients treated with regimens including Ambisome.