We report the results of an external quality control program, including 17 Italian centers involved in the care of patients infected by HIV, to evaluate CD4 T cell count proficiency and reproducibility. The centers received two commercial stabilized blood preparations, one with "normal" and one with "low" CD4 T cell content. The centers were asked to process the samples two times, 1 week apart, with the same procedure used for samples from HIV patients. Most centers showed a good performance of CD4 frequency and absolute count determinations. In particular, the "low" sample was correctly analyzed by all centers; only two underestimated the "normal" sample CD4 frequency, and only one underestimated the CD4 absolute count by >100 CD4 cells/microl. Overall, our data suggest that most Italian laboratories provide reliable and reproducible results in evaluating CD4 T cells in HIV(+) samples.