To evaluate the effect of antileukemic chemotherapy administered at diagnosis on the survival of patients with isolated chloroma. Retrospective review of locally identified patients and analysis of cases from the medical literature. The records of all patients with isolated chloroma identified at three teaching hospitals in Toronto between 1980 and 1994 were reviewed. A MEDLINE search was done to identify all cases of isolated chloroma reported in the English-language medical literature. Patients with a previous known hematologic disorder were excluded. The effect of therapy on 1) the interval between diagnosis of chloroma and diagnosis of acute myeloid leukemia and 2) survival was determined. 7 local patients and 83 published cases were identified, for a total of 90 evaluable patients. For the entire group, the median time to the diagnosis of acute myeloid leukemia was 9 months, and median survival was 22 months. Chemotherapy was administered to 49 patients (54%) at diagnosis of chloroma. Significantly fewer patients treated with chemotherapy subsequently developed acute myeloid leukemia (41% compared with 71%; P = 0.001). Survival was longer in patients treated with chemotherapy (> 50% alive with a median follow-up of 25 months compared with a median survival of 13 months for those initially untreated; P = 0.001). Multivariate analysis showed that neither local radiotherapy nor surgery had an effect on survival. Administration of antileukemic chemotherapy at diagnosis of chloroma is associated with a significantly lower probability of developing acute myeloid leukemia and with longer survival.