Patients who undergo autologous hematopoietic stem cell transplantation (aHCT) for treatment of a relapsed or refractory lymphoma are at risk of developing therapy related-myelo-dysplasia/acute myeloid leukemia (t-MDS/AML). Part of the risk likely resides in inherent interindividual differences in their DNA repair capacity (DRC), which is thought to influence the effect chemotherapeutic treatments have on the patient's stem cells prior to aHCT. Measuring DRC involves identifying small differences in repair proficiency among individuals. Initially, we investigated the cell model in healthy individuals (primary lymphocytes and/or lymphoblastoid cell lines) that would be appropriate to measure genetically determined DRC using host-cell reactivation assays. We present evidence that interindividual differences in DRC double-strand break repair (by non-homologous end-joining [NHEJ] or single-strand annealing [SSA]) are better preserved in non-induced primary lymphocytes. In contrast, lymphocytes induced to proliferate are required to assay base excision (BER) or nucleotide excision repair (NER). We established that both NHEJ and SSA DRCs in lymphocytes of healthy individuals were inversely correlated with the age of the donor, indicating that DSB repair in lymphocytes is likely not a constant feature but rather something that decreases with age (~0.37% NHEJ DRC/year). To investigate the predictive value of pre-aHCT DRC on outcome in patients, we then applied the optimized assays to the analysis of primary lymphocytes from lymphoma patients and found that individuals who later developed t-MDS/AML (cases) were indistinguishable in their DRC from controls who never developed t-MDS/AML. However, when DRC was investigated shortly after aHCT in the same individuals (21.6 months later on average), aHCT patients (both cases and controls) showed a significant decrease in DSB repair measurements. The average decrease of 6.9% in NHEJ DRC observed among aHCT patients was much higher than the 0.65% predicted for such a short time frame, based on ageing results for healthy individuals.