Ocean tide loading (OTL) displacements of eight principal constituents at 12 sites in Hong Kong were determined using more than eight years of continuous GPS observations. Parameters of the OTL displacements were estimated using daily GPS solutions obtained with precise point positioning (PPP) technique. The results were compared with predictions from seven latest global ocean tide models. Gravity measurements of OTL in Hong Kong were also used to validate the GPS results. The study shows that the results from both the GPS and gravity measurements agree best with the GOT4.7 and NAO99b models, when the K1 and K2 constituents are excluded. The agreements between the GPS and the model estimates are generally at sub-millimeter level in both the horizontal and vertical directions, except for S2, K2 and K1 constituents that have relatively larger errors. After removing the systematic biases between the GPS and the model estimates, the misfits of M2, S2, N2, O1, P1 and Q1 at all sites are within 0.5 and 1.0 mm in the horizontal and the vertical directions, respectively, whereas K1 and K2 show relatively larger misfits of up to 2.3 mm. Both the GPS and the gravity S2 estimates have large biases with unknown reasons when compared with the modeled values, which needs to be further investigated. The study demonstrates that GPS is capable of estimating the OTL displacements with the same accuracy as the model predictions, especially for coastal areas. Parameters of the OTL displacements were estimated using daily GPS solutions obtained with precise point positioning (PPP) technique. The results were compared with predictions from seven latest global ocean tide models. Gravity measurements of OTL in Hong Kong were also used validate the GPS results. The study shows that the results from both the GPS and gravity measurements agree best with the GOT 4.7 and NAO99b models, when the K1 and K2 constituents are excluded. The agreements between the GPS and the model estimates are generally at sub-millimeter level both in the horizontal and vertical directions, except for S2, K2 and K1 constituents that have relatively large errors After removing the systematic biases between the GPS and the model estimates, the misfits of M2, S2, N2, O1, P1 and Q1 at all sites are wit hin 0.5 and 1.0 mm in the horizontal and the vertical directions, respectively, while K1 and K2 show relatively larger misfits of up to 2.3 mm. Both the GPS and the gravity S2 estimates have large biases with unknown reasons when compared with the modeled values, which needs to be further investigated. The study demonstrates that GPS is capable of estimating the OTL displacements with the same accuracy as the model predictions, especially for coastal areas.