At present, 60% to 90% of the patients (pts) with OVCA demonstrate objective responses to first-line platinum (Pt)-based combination chemotherapy (CT). Serial CA 125 measurements reflect the clinical course of the disease in these circumstances correctly in 89% of the cases (Neth J Med 40:36, 1992). In pts treated with paclitaxel (Taxol ® , T) after Pt-based CT this correlation is poor (Proc ECCO-VII; 133, 1993). The question can be raised whether this poor correlation is a general phenomenon in relapsed OVCA pts or whether this is related to the use of T.to determine clinical response and CA 125 response with Pt-based CT after T treatment and to assess the correlation of serum CA 125 levels with the clinical course with both treatments, i.e. with T and with Pt-based CT after T.Doubling or halving of CA 125 levels were considered to be a significant increase or decrease. For the relationship with the clinical course we used the Spearman rank correlation.So far, 18 pts (with 6/18 clinical responses and 15/16 marker responses on T) were pre-treated with a Pt-based CT, 9 in 3rd-line, 6 in 4th-line, and 3 in 5th-line. 7 Pts responded (39%; 2 CR, 5 PR) according to WHO criteria and 8 of 15 evaluable pts (53%) had a CA 125 response. The correlation between changes in CA 125 levels and clinical course was poor for T (correlation 0.27; P=0.31), but significant for Pt-based CT (correlation 0.67; P=0.0066).1) Pt-compounds and T are not cross resistant 2) the poor correlation between changes in serum CA 125 levels and the clinical course of the disease seems to be specific for treatment with T.