It is known that the number of transplanted cells has a significant impact on the outcome after SCT. We identify issues that cannot be addressed by conventional analysis of clinical trials and ask whether it is possible to develop a refined analysis to conclude about the outcome of individual patients given clinical trial results. To accomplish this, we propose an interdisciplinary approach based on mathematical modeling. We devise and calibrate a mathematical model of short-term reconstitution and simulate treatment of large patient groups with random interindividual variation. Relating model simulations to clinical data allows quantifying the effect of transplant size on reconstitution time in the terms of patient populations and individual patients. The model confirms the existence of lower bounds on cell dose necessary for secure and efficient reconstitution but suggests that for some patient subpopulations higher thresholds might be appropriate. Simulations demonstrate that relative time gain because of increased cell dose is an 'interpersonally stable' parameter, in other words that slowly engrafting patients profit more from transplant enlargements than average cases. We propose a simple mathematical formula to approximate the effect of changes of transplant size on reconstitution time.