The last decade has seen widespread enthusiasm around the development of novel epigenetic therapies for the treatment of many diseases, including cancer. BET-bromodomain inhibitors are amongst the most exciting, with promising pre-clinical data paving the way for clinical trials. In these trials, some patients achieve a moderately durable partial or complete response; however, there remains much to be learnt about these therapies to further optimize their efficacy. The current generation of BET inhibitors do not discriminate between the two tandem bromodomains of the BET family members and therefore tolerability and clinical outcomes may be limited due to a lack of specificity for the individual bromodomains and/or BET family members. In collaboration with GSK, we have characterised domain-selective BET inhibitors with >200-fold selectivity for either the first bromodomain (BD1) or second bromodomain (BD2) of all the BET family proteins. In vitro and in vivo functional assays demonstrate that the BD1-selective compound phenocopies pan-BET (I-BET, JQ1) inhibitors, while inhibition of BD2 has relatively modest effects in cancer cells. Remarkably, and in contrast to their role in cancer cells, we find that BD2 selective compounds have a significant role in modulating immune-inflammatory processes where its phenotypic effects are similar to the non-selective BET inhibitors. We will present a detailed analysis of these phenotypic responses and provide a molecular understanding of the different requirements for the individual bromodomains during cell context dependent gene expression. Taken together, we describe specific functions for the individual BET family bromodomains in the regulation of gene expression, raising the possibility of specifically targeting cancer associated programs (via BD1) or immunogenic responses (via BD2) to improve the efficacy and tolerability of current non-selective BET inhibitors.