The Large Underground Xenon experiment (LUX) searches for dark matter using a dual-phase xenon detector. LUX uses a custom-developed trigger system for event selection. In this paper, the trigger efficiency, which is defined as the probability that an event of interest is selected for offline analysis, is studied using raw data obtained from both electron recoil (ER) and nuclear recoil (NR) calibrations. The measured efficiency exceeds 98% at a pulse area of 90 detected photons, which is well below the WIMP analysis threshold on the S2 pulse area. The efficiency also exceeds 98% at recoil energies of 0.2 keV and above for ER, and 1.3 keV and above for NR. The measured trigger efficiency varies between 99% and 100% over the fiducial volume of the detector.