A phase I/II trial was designed to evaluate the safety and efficacy of adding bortezomib to standard BEAM (BCNU, etoposide, cytarabine, melphalan) and autologous hematopoietic stem cell transplantation (ASCT). Eligible patients had relapsed/refractory indolent or transformed non-Hodgkin lymphoma or mantle cell lymphoma (MCL) that was relapsed/refractory or in first partial (PR) or complete remission (CR). Patients received bortezomib on days -11, -8, -5, and -2 before ASCT. Phase I had 4 dose cohorts (.8, 1, 1.3, and 1.5 mg/m(2)) and 3 patients were accrued to each. Any nonhematological ASCT-related toxicity >2 on the Bearman scale occurring between day -11 and engraftment defined the maximum tolerated dose (MTD). After the MTD has been reached, another 20 patients were enrolled at this dose to determine a preliminary overall response rate (ORR). Patients who were in CR or PR at day +100 were considered responders. The study enrolled 42 patients through August 14, 2009. The median age was 58 (range, 34 to 73) years, with 33 males and 9 females. The most common diagnoses were MCL (23 patients) and follicular lymphoma (7 patients). The median number of prior therapies was 1 (range, 0 to 6). The median follow-up was 4.88 (range, 1.07 to 6.98) years. Thirteen patients were treated in phase I and 29 patients were treated in phase II. The MTD was initially determined to be 1.5 mg/m(2) but it was later decreased to 1 mg/m(2) because of excessive gastrointestinal toxicity and peripheral neuropathy. The ORR was 95% at 100 days and 87% at 1 year. For all 38 evaluable patients at 1 year, responses were CR 84%, PR 1%, and progressive disease 13%. Progression-free survival (PFS) was 83% (95% CI, 68% to 92%) at 1 year, and 32% (15% to 51%) at 5 years. Overall survival (OS) was 91% (95% CI, 79% to 96%) at 1 year and 67% (50% to 79%) at 5 years. The most common National Cancer Institute grade 3 toxicities were neutropenic fever (59%), anorexia (21%), peripheral neuropathy (19%), orthostatic hypotension/vasovagal syncope (16%), and 1 patient failed to engraft. Compared with 26 MCL in CR1 historic controls treated with BEAM and ASCT, PFS was 85% and 43% for the BEAM group versus 87% and 57% for those who received bortezomib in addition to standard BEAM (V-BEAM) at 1 and 5 years, respectively (log-rank P = .37). OS was 88% and 50% for the BEAM group versus 96% and 72% for V-BEAM at 1 and 5 years, respectively (log-rank P = .78). In conclusion, V-BEAM and ASCT is feasible. The toxicities were manageable and we did not observe any treatment-related mortalities; however, we did observe an excess of autonomic dysfunction and ileus, which is concerning for overlapping toxicity with BEAM conditioning. Determining relative efficacy of V-BEAM compared to BEAM would require a randomized trial.