We present the case of a 49-year-old Hispanic man with HIV. The patient was diagnosed HIV positive in 1984, the infection most likely resulting from homosexual contact. He was also hepatitis A and hepatitis B positive, was hepatitis C negative and had negative syphilis serology. His total and fractionated bilirubin levels were normal and there was no history of Gilbert's disease in his medical record. In 2001, he suffered an episode of bacterial pneumonia and in 2002 he reached his CD4 count nadir, with 24 cells/mm3 and an HIV-RNA plasma viral load (HIV-RNA pVL) of more than 600,000 copies/ml. In 2003, he was diagnosed and treated for an oral Kaposi's sarcoma. As a flight attendant, he had a frenetic lifestyle and was treated for HIV infection in different countries. As a result, his medication history included several HAART combinations: (1) tenofovir disoproxil + didanosine + efavirenz (TDF + ddI + EFV); (2) tenofovir disoproxil + zidovudine + lopinavir/ritonavir (TDF + AZT + LPV/r); (3) stavudine + didanosine + lopinavir/ritonavir (d4T+ ddI + LPV/r); (4) stavudine + didanosine + efavirenz (d4T+ ddI + EFV); and (5) didanosine 250 mg once daily + tenofovir disoproxil + efavirenz (ddI 250 + TDF + EFV). The reasons for switching from one regimen to another between 2002 and January 2005 included virological failure, diarrhea and the development of lipodystrophy. When we first evaluated the patient in January 2005, his CD4 level was 208 cells/mm3 and his viral load was approximately 8000 copies/ml. Genotype testing showed a 190Q mutation related to EFV and a 65R mutation related to TDF. In the protease domain, only a 63P polymorphism was evident. A new treatment regimen was started of ddI (400 mg once daily), d4T (40 mg twice daily), and ritonavir-boosted saquinavir (SQV/r), 1000 mg/100 mg twice daily. In September 2005, his HIV-RNA pVL was undetectable and CD4 cell count had risen to 232 cells/mm3. At this time, his total bilirubin concentration was 1.40 mg/dl (unconjugated bilirubin 1 mg/dl; conjugated bilirubin 0.40 mg/dl). A few weeks later, the patient returned to his home country in South America and was examined again in March 2006, during which his viro-immunological control was close to that obtained in Italy, six months previously. In May 2006, he returned to our center complaining of colicky abdominal pain and diarrhea with up to eight discharges of loose stools per day. Scleral icterus was noted on physical examination. After excluding drug toxicities, viral hepatitis and alcohol abuse, it became evident that he had been switched from SQV/r to ATV/r while in his native country. Laboratory data showed grade 3 hyperbilirubinemia (total bilirubin 5.4 mg/dl; unconjugated bilirubin 5.1 mg/dl; conjugated bilirubin 0.3 mg/dl) while his HIV-RNA pVL and CD4 count remained basically unchanged. His trough ATV plasma concentration was measured by high performance liquid chromatography (HPLC) and found to be 5.2 μg/ml, around ten times the therapeutic level of 0.15-0.85 μg/ml []. The accuracy of drug intake and blood sampling were checked and confirmed as correct. The patient sought medical care in Milan due to the acute onset of vomiting and diarrhea in May 2006, one week after being examined in our institution. A serum total bilirubin concentration of 7.6 mg/dl (unconjugated bilirubin 7 mg/dl; conjugated bilirubin 0.6 mg/dl) was noted but transaminase levels were within the normal range, as were gamma-GT and alkaline phosphatase. He was discharged from the emergency room with his gastrointestinal symptoms alleviated. In June 2006, the patient returned to our clinic still showing mild scleral icterus but was otherwise asymptomatic. At this stage he presented a note from his medical practitioner from South America that stated he had been taking 300 mg of ATV boosted by 200 mg of ritonavir (RTV). Hence, an incorrect RTV dosage had been taken since May 2006. A further HPLC test in June 2006 confirmed a high level of ATV Ctrough (5.4 μg/ml). Interestingly, the total bilirubin concentration had spontaneously decreased to around 3 mg/dl. The RTV boosting dosage was reduced to 100 mg, resulting in a decrease of scleral icterus and a total bilirubin concentration of 4.3 mg/ml (unconjugated bilirubin 3.9 mg/dl; conjugated bilirubin 0.4 mg/dl). His ATV Ctrough levels also fell, but were still above the therapeutic range. The patient had subsequent monthly assessments. In July 2006, he presented with mild scleral icterus, a CD4 cell count > 300 cells/mm3, and an undetectable viral load. However, his total bilirubin and ATV Ctrough were unchanged. In August 2006, the patient's jaundice disappeared, his serum total bilirubin concentration was markedly reduced, and his CD4 cell count was unchanged, though levels of HIV-RNA pVL had increased to approximately 200 copies/ml. His ATV Ctrough level was below detection limit (0.1 μg/ml), despite his adherence to the therapy. Thus, both the total bilirubin and the ATV Ctrough decreased after reduction of the RTV boosting dosage. At his last clinic visit in August 2006, the patient's total bilirubin concentration was normal, but the ATV Ctrough was below the therapeutic range. The patient did not return to our clinic and was therefore lost to follow-up.