Purpose: Double protease inhibitor (PI) boosting is being explored as a new strategy in salvage antiretroviral (ARV) therapy. However, if a negative drug interaction leads to decreased drug levels of either or both PIs, double PI boosting could lead to decreased virologic response. A negative drug interaction has been described between amprenavir (APV) and lopinavir/ritonavir (LPV/r). This observational cohort study assessed the virologic impact of the addition of APV to a salvage ARV regimen, which also contains LPV/r, compared to a regimen containing LPV/r alone. Method: Patients initiated on a salvage ARV regimen that included LPV/r obtained from the expanded access program in Toronto, Canada, were evaluated. APV (600-1,200 mg bid) was added at the discretion of the treating physician. Results: Using multivariate Cox proportional hazards models, we found that the addition of APV to a LPV/r-containing salvage regimen was not significantly associated with time to virologic suppression (< 50 copies/mL; adjusted hazard ratio [HR] = 0.75, p = .12) or with time to virologic rebound (adjusted HR = 1.46, p = .34). Those patients who received higher doses of APV had an increased chance of virologic suppression (p = .03). In a subset of 27 patients, the median LPV Ctrough was significantly lower in patients receiving APV (p = .04), and the median APV Ctrough was reduced compared to reported controls. Conclusion: Our data do not support an additional benefit in virologic reduction of double boosting with APV and LPV/r relative to LPV/r alone in salvage ARV therapy. Our study's limitations include its retrospective nature and the imbalance between the two groups potentially confounding the results. Although these factors were adjusted for in the multivariate analysis, a prospective randomized controlled trial is warranted to confirm our findings.