The solution of a linear model of barotropic M2 tides in Bass Strait is not consistent with new high quality tide gage and current meter data. This is attributed to the method used to estimate the forcing due to incoming radiation at the open boundaries, using only historical tide data. The new data is assimilated into the model using a generalized inverse theory described by Bennett and McIntosh. The analysis establishes the consistency of the new data with the model dynamics of linearized shallow water waves, and uses those dynamics to interpolate between the new data. The dependence of the solution upon the estimated boundary forcing is reduced. Relaxation of the coastal rigid boundary condition, within estimated model interpolation errors, has a significant effect on the solution. Abstract The solution of a linear model of barotropic M2 tides in Bass Strait is not consistent with new high quality tide gage and current meter data. This is attributed to the method used to estimate the forcing due to incoming radiation at the open boundaries, using only historical tide data. The new data is assimilated into the model using a generalized inverse theory described by Bennett and McIntosh. The analysis establishes the consistency of the new data with the model dynamics of linearized shallow water waves, and uses those dynamics to interpolate between the new data. The dependence of the solution upon the estimated boundary forcing is reduced. Relaxation of the coastal rigid boundary condition, within estimated model interpolation errors, has a significant effect on the solution.