Our main goal was to evaluate the CD34+ dose in patients undergoing haemotopoietic stem celltransplantation and its results in terms of recovery of neutrophile and platelet counts, transfusion requirements, days of fever, antibiotic requirements and length of hospital stay. We studied 38 consecutive patients with haematological malignancies transplanted at our Department, from Feb. 96 through Sept. 98. The CD34+ cell quantification technique was standardized, using a modification of the ISAGHE 96 protocol. Patients were sorted into three groups according to the CD34+ count administered: a) between 3 and 5 x 10(6) cells/kg; b) between 5 and 10 x 10(6) cells/kg; c) > 10 x 10(6) CD34+ cells/kg. As a secondary end point, results were assessed according to the number of aphereses required to arrive at the target count of CD34+, separating those patients that required only 1 or 2 aphereses versus those requiring 3 or more. Finally, an analysis was made of the results of transplantation comparing the different sources of stem cells (PBSC versus PBSC + B.M.). The best results were obtained in the group with cells between 3 and 5 x 10(6) CD34+. No statistically significant advantages were found in the group with cells over 5. The supra-optimal dose of more 10 x 10(6) would yield no additional beneficial results, while they can imply a greater infusion of residual tumor cells. The number of aphereses had no impact on engraftment. Results obtained with PBSC transplants were better than those with BM+PBSC in terms of neutrophile and platelet recovery. The number of CD34+ cells remains the main element in stem cell transplantation to evaluate the haematopoietic recovery after engraftment. Minimum and optimum yields remain unclear. Centers should establish their own optimal dose based on local methodologies and outcomes, maximizing costs and benefits.