Both CD34 (cluster of differentiation 34) and the more recently described CD133 are markers of primitive stem cells with haematopoietic repopulating ability. Most transplanting centres use a minimum number of CD34+ cells as the requirement for a transplant and consider this a predictor of haematopoietic engraftment. However, transplanted CD34+ cell dose does not always give a close correlation with time to engraftment nor explain delayed engraftment in some patients. We have retrospectively evaluated the potential of measuring viable CD133+ cell numbers in the autograft as an alternative predictor of haematological engraftment after autologous stem-cell transplantation in a cohort of patients with multiple myeloma (MM). We found an average 32% loss of viability of CD34+ cells in the post-thaw sample compared with the fresh sample. Of the original estimated CD34+ cell numbers transplanted per kg, 43% of the thawed samples were double positive for CD34+/CD133+. In this patient group, the CD34+/CD133+ subset gave the closest statistical correlation with time to neutrophil engraftment (p < 0.05), particularly for patients given above median (1.8 x 10(6)/kg) dose of the double-positive cells. The CD34+/CD133+ population was the only parameter to give a significant correlation with white cell engraftment in this patient cohort (p < 0.05). There was no significant correlation between CD34+, viable CD34+ or viable CD34+/CD133+ cells/kilogram with platelet engraftment. Determination of viable CD34+/CD133+ progenitor cell dose in the autograft may be a useful tool to predict neutrophil recovery after autologous transplantation than conventional assessment of CD34+ numbers. These results warrant further investigation of the role of CD133 in haematopoietic engraftment.