OBJECTIVE Chronic myeloid leukemia (CML). METHODS CML patients confirmed at the OAUTHC, Ile-Ife between June 1986 and December 1999 were studied prospectively until death or the last visit to the clinic. Stages of the disease at diagnosis, presenting clinical features and laboratory parameters, as well as the drug history were noted for each patient. Survival of each patient was computed from the date of diagnosis until the date of death. Kaplan-Meier statistical technique was used for analysis of survival. Factors were compared with log rank test. Student's t-test was used for comparing means of presenting clinical and laboratory parameters between groups. P-value of <0.05 was considered significant. SPSS for windows Release 10 (Real Stats, Real Easy, 1989-1999) was used for all calculations. CONCLUSIONS There were 42 patients (27 males and 15 females, ratio of 1.8:1) with ages ranging from 12-74 (median=38.0) years. Philadelphia chromosome was positive in the five patients (12.0%) that underwent cytogenetic analysis. Thirty-three of the patients (78.6%) presented in treatable chronic phase and the other 9 (21.4%) were seen in advanced stages (7 in accelerated & 2 in blastic phase). The median survival of patients that presented in chronic phase was 31.7 months compared to 0.16 months in patients presenting in advanced stages, the difference was statistically significant (log rank=7.8, p-value=0.005). Significant positive correlation was obtained between spleen size and total white cell count at diagnosis (r=0.36, p=0.02). Univariate regression analysis showed negative relationship between survival and ages of patients at diagnosis, haematocrit value, spleen and liver sizes, and blast count. The relationships were statistically significant for only liver and spleen (b=-0.63, p=0.004 and b=-0.51, p=0.026, respectively). Regression of both the liver and spleen with survival, still confirms the significant negative relationship between hepatomegaly and survival.