RNA viruses like HIV and HCV have an extraordinary evolutionary potential to escape from both immune pressures and targeted drug therapies. In HIV infections, the emergence of drug resistant strains is of particular interest, as it complicates the choice of an optimal follow-up regimen. A series of bioinformatics tools for predicting drug resistance were previously developed to support physicians in this task. A new method is proposed that captures the order of occurrence of drug-resistant mutations and can be applied to serially-sampled viral sequence data from patients taking antiretroviral drugs. The new phylogenetic approach reduces a serial evolutionary tree inferred by the Sliding MinPD program [12] to a set of mutational pathways of drug resistance. The method is applied to data from an HIV-1 clinical study of the reverse transcriptase inhibitor, Efavirenz. This approach can effectively identify mutational pathways by considering all available information and the statistical support for each prediction.