The duration of the Science Mission is now sufficient to resolve and map baroclinic tides at the dominant semidiurnal frequencies of M2 and S2. New estimates are derived from SWOT using a remove/restore strategy to identify features of the baroclinic tides not resolved in older baroclinic tide models. Tidal predictions based on the new estimates typically explain from 3% to 20% more SSH variance than HRET8.1 as judged against the CryoSat-2 observations, which are used for validation. Based on regions examined thus far, the largest accuracy gains are achieved in areas with complex coastlines and large-amplitude internal tides, such as the Banda Sea. There is evidence from parts of the North Pacific that SWOT-based tidal estimates capture interannual variability of baroclinic tides, since the predictions are less accurate for CryoSat-2 data collected in years prior to SWOT launch. Development of the tidal estimates is an ongoing exercise, since results are sensitive to residual long-wavelength errors and data flagging/selection, aspects of the SWOT data which are evolving in different product releases. Results also depend on assumptions used to smooth SWOT-derived harmonic constants and interpolate them into data gaps between swaths.