The impact of treatment-related factors on inhibitor development in previously untreated patients (PUPs) with haemophilia A is still debated. We present the results of a collaborative, individual patient data meta-analytic project. Eligible data sources were published cohorts of PUPs for which patient-level data were available. The exposures of interest were factor (F)VIII type (recombinant [rFVIII] vs plasma-derived [pdFVIII]) and treatment intensity (≥ vs < 150 IU/kg/week) at first treatment. Family history of inhibitors, F8 mutations, age, treatment regimen (on-demand vs prophylaxis), secular trend and surgery were analysed as putative confounders using different statistical approaches (multivariable Cox regression, propensity score analyses, CART). Analyses accounted for the multi-centre origin of the data. We included 761 consecutive, unselected PUPs with moderate to severe haemophilia A from 10 centres in Egypt, Germany, Israel and Italy. A total of 27 % of patients developed inhibitors; 40 % and 22 % of patients treated with rFVIII and pdFVIII (unadjusted HR 2.2, 95 % CI 1.6–2.9), respectively; 51 % and 24 % of patients receiving high-and low-intensity treatment (unadjusted HR 2.9, 95 % CI 2.0–4.2), respectively. In adjusted analyses, only treatment intensity remained an independent predictor; the effect of FVIII type was largely due to confounding, but with a significant interaction between FVIII type and treatment intensity. This patient-level meta-analysis confirms, across different statistical approaches, that high-intensity treatment is a strong risk factor for inhibitor development. The possible role of FVIII type in subgroups is suggested by the test for interactions but could not be proven because of the limited subgroups sample sizes.