In children and adolescents, prophylactic application of G-CSF after myelosuppressive chemotherapy reduces the duration of neutropenia, of hospitalization, and of parenteral antibiotic treatment. In acute lymphoblastic leukemia, non-Hodgkin lymphoma, and solid tumors, G-CSF support allows dose intensification of chemotherapy. On the other hand, each G-CSF injection causes pain and costs. We set up the hypothesis that besides strategies as restricted indications, delayed start, lower doses, and stringent rules for discontinuation, individualized timing of blood count might optimize G-CSF prophylaxis. We randomized 64 cycles of G-CSF prophylaxis in eight children and adolescents being treated for acute lymphoblastic leukemia or solid tumors to standard twice a week or to individually timed blood counts. Primary study endpoints were the numbers of G-CSF doses and of blood counts, and the total costs of G-CSF support. Per cycle, individual timing of blood count resulted in a median of one G-CSF injection fewer [estimated population median (EPM) 1.40, 95% confidence interval (CI) 0.57-2.20] and one blood count fewer (EPM 1.00, 95% CI 0.74-1.33). The total costs of G-CSF support and of blood counts were thereby reduced by a median of US $ 152 per cycle (EPM 191, 95% CI 97-318). The results of this study suggest that individual timing of blood counts during prophylactic G-CSF support in children and adolescents undergoing chemotherapy for malignant disease can significantly reduce the number of injections and blood counts performed, thus resulting in less pain and lower costs. These findings need confirmation in a larger randomized trial.