We have analysed 10 ks of XRT data for GRB 100728A (Cannizzo et al. GCN 11004), from 66 s to 23.4 ks after the BAT trigger. The data comprise 1.3 ks in Windowed Timing (WT) mode (the first 9 s were taken while Swift was slewing) with the remainder in Photon Counting (PC) mode. The enhanced XRT position for this burst was given by Beardmore et al. (GCN. Circ 11005). The light curve can be modeled with a broken power-law decay, with a series of flares superimposed during the first kilosecond. The underlying decay begins with a decay index of 1.6 (+/-0.1), breaking at T+410 sec to a decay index of 1.07 (+0.03, -0.04). The decay then breaks again at T+11600 sec to a final decay of 1.82 (+/-0.2). Nine discrete flares can clearly be identified, with peak times of T+ 88, 111, 189, 252, 296, 383, 445, 516, 659 sec. These all occurred while the XRT was in Windowed Timing mode. A spectrum formed from the WT mode data can be fitted with an absorbed power-law with a photon spectral index of 2.022 (+/-0.022). The best-fitting absorption column is 3.69 (+0.09, -0.08) x 10^21 cm^-2, in excess of the Galactic value of 1.0 x 10^21 cm^-2 (Kalberla et al. 2005). The PC mode spectrum has a photon index of 2.05 (+/-0.09) and a best-fitting absorption column of 3.7 (+/-0.3) x 10^21 cm^-2. The counts to observed (unabsorbed) 0.3-10 keV flux conversion factor deduced from this spectrum is 4.5 x 10^-11 (7.6 x 10^-11) erg cm^-2 count^-1. If the light curve continues to decay with a power-law decay index of 1.82, the count rate at T+24 hours will be 0.024 count s^-1, corresponding to an observed (unabsorbed) 0.3-10 keV flux of 1.1 x 10^-12 (1.8 x 10^-12) erg cm^-2 s^-1. The results of the XRT-team automatic analysis are available at http://www.swift.ac.uk/xrt_products/00430151. This circular is an official product of the Swift-XRT team.