Evidence for intravenous immunoglobulin (IVIG) in recurrent pregnancy loss remains inconsistent. This individual patient data meta-analysis aimed to explore possible treatment-effect modification by the number of previous pregnancy losses across six randomized trials. One-stage logistic regression included treatment, number of previous losses, their interaction, maternal age, and study-specific intercepts. The estimated live birth odds ratio for each additional loss was 0.6 in the placebo group and significantly higher 0.9 in the IVIG group (ratio of odds ratios 1.50, 95% CI 1.06-2.18). These exploratory findings suggest that the effect of IVIG on live birth odds may increase with previous loss burden.