In the present review we analyse the current knowledge about the growth properties of AML progenitor cells and their relationship with other clinico-biological characteristics of the disease. Leukaemic colony forming unit L-CFU is considered to be the clonogenic cell in AML and more immature than the blast cell population. Our studies have shown that in leukaemic hematopoiesis colony forming cells can exist among both cell fractions CD34+ and CD34-. Optimal "in vitro" proliferation of L-CFU is dependent upon the addition of exogenous growth factors. However, it has been observed that leukaemic progenitor cells frequently display a certain degree of autonomous proliferation. In order to quantify the "in vitro" behaviour of L-CFU, we have explored 3 parameters: 1) plating efficiency (PE); 2) autonomous growth (AG); and 3) autonomous proliferative index (API) which was calculated as AG divided by PE and we have correlated them with other clinico-biological data. According to the FAB classification we could observe that patients with M3 subtype showed an higher PE than other AML subgroups and a significantly lower API. Regarding CD34 expression we observed that AG was enhanced in CD34+ cases and also in those showing a higher rh123 elimination. In order to determine whether PE could condition clinical evolution, we analysed this parameter in a large series of patients but failed to demonstrate any relationship. By contrast, we observed that patients who displayed a higher API showed a shorter survival than patients with lower API (18% vs 48% surviving at 3 years). We have also shown that abnormalities in the CFU-GM growth pattern could be associated with risk the of relapse in AML patients; a switch from normal to abnormal "in vitro" growth should alert us. But for the assessment of the real value of these analyses sequential follow-up studies are mandatory. In summary, cell culture studies contribute not only to a better understanding of leukaemic hematopoiesis but may also contribute to better disease monitoring.