Over the last decade we have successfully evaluated the use of a human skin explant assay for predicting acute GVHD in HLA-matched sibling transplants. In the present study, we modified GVHD prophylaxis on an individual patient basis depending on the GVHD outcome predicted by the skin explant model. We have summarised our previous data describing how the skin explant assay results correctly predict GVHD occurrence and severity in 45/56 patients (80%); P< 0.0001, chi2 19.97, df = 1. In a further cohort of 19 patients, all were predicted to develop grade II or above GVHD. These patients were given increased GVHD prophylaxis with the addition of methotrexate and a significant reduction in the expected incidence of GVHD was observed (P = 0.02; chi2 7.7, df = 1; Fisher exact test P = 0.04). The results from these studies suggest that modifying GVHD prophylaxis, based on skin explant assay results, may reduce the expected incidence and severity of GVHD. We suggest that the technique might be used for selective GVHD prophylaxis in T cell non-depleted HLA matched sibling transplants.