The bromodomain and extra-terminal domain (BET) proteins are epigenetic readers, regulating transcription via two highly homologous tandem bromodomains, BD1 and BD2. Clinical development of nonselective pan-BD BET inhibitors has been challenging, partly due to dose-limiting side effects such as thrombocytopenia. This has prompted the push for domain-selective BET inhibitors to achieve a more favorable therapeutic window. We report a structure-guided drug design campaign that led to the development of a potent BD1-selective BET inhibitor, 33 (XL-126), with a Kd of 8.9 nM and 185-fold BD1/BD2 selectivity. The high selectivity was first assayed by SPR, validated by a secondary time-resolved fluorescence energy transfer assay, and further corroborated by BROMOscan (∼57-373 fold selectivity). The cocrystal of 33 with BRD4 BD1 and BD2 demonstrates the source of selectivity: repulsion with His437 and lost binding with the leucine clamp. Notably, the BD1 selectivity of BET inhibitor 33 leads to both the preservation of platelets and potent anti-inflammatory efficacy.