The results of the IFM 95-01 trial comparing MP to Dexamethasone-based regimens confirmed that the MP treatment remained the reference for patients (pts) older than 65 years with multiple myeloma (MM) (T. Facon et al., Blood 2005, in press). In May 2000, the IFM initiated a new trial, IFM 99-06, for pts aged 65–75 years, comparing MP (12 courses at 6 weeks intervals) to MP-THAL (MP plus THAL at the maximum tolerated dose, but d 400 mg/day, stopped at the end of MP) and a MEL100-based treatment (intermediate-dose MEL). The third arm’s schedule was VADx2, cyclophosphamide 3g/m2 for stem cell collection and 2 courses of MEL100 mg/m2. IFM99-06 was planned to enroll 476 pts for evaluation, whose treatment allocation followed a 3(MP)-2(MP-THAL)-2(MEL100) randomization scheme. The primary end-point was overall survival (OS), with two primary comparisons, MP vs MP-THAL and MP vs MEL100, and a secondary one, MP-THAL vs. MEL100. Secondary end-points were response to treatment and progression-free survival (PFS). Time-to-event curves were compared using the Cox model on an intent-to-treat basis and results expressed through hazard ratio (RR) with 95% confidence interval (95%CI). Two interim analyses were planned, according to Peto’s rule, after the enrollment of 200 and 350 pts, and reviewed by a Data Safety Monitoring Board (DSMB) independent of IFM. At the second interim analysis, the DSMB recommended the trial be pursued as planned, but suggested that a third interim analysis be performed with a date of point on May 1, 2005. At this time, 436 pts had been enrolled, 191, 124 and 121 in MP, MP-THAL and MEL100 groups, respectively. There were no differences in the distributions of pre-treatment characteristics between treatment groups, except for serum calcium level (P=0.02). The median (se) follow-up time was 32.2 (1.8) months (mo.), similar across groups. Median (se) PFS times were 17.1 (1.4), 27.6 (3.6) and 19.0 (1.2) mo. in MP, MP-THAL and MEL100 groups, respectively. The PFS time was significantly longer in the MP-THAL group than in the MP group (hazard ratio estimate, RR=2.4, 95% CI=1.8–3.3, P