In 1969, prior to the discovery of the subglacial Lake Vostok, an Askania Gs-11 gravimeter was operated at Vostok Station (78.466 degrees S, 106.832 degrees E; 3478 m asl) to observe tidal gravity variations. To gain a better understanding of the lake's tidal dynamics, we reanalyzed these data using a Bayesian Tidal Analysis Program Grouping method (BAYTAP-G and -L programs). The obtained phase leads for the semidiurnal waves M2 (6.6 +/- 2.1 degrees) and S2 (10.1 +/- 4.2 degrees) are more pronounced than those of the diurnal waves, among which the largest phase lead (for K1) was 5.0 +/- 0.5 degrees. The obtained delta factor for M2 was 0.890 +/- 0.032, significantly less than the theoretical value of 1.16. For three global ocean tide models (NAO99b, FES2004, and TPXO6.2), the estimated load tides on waves Q1, O1, P1, K1, M2, and S2 range from 0.1-0.2 mu Gal (Q1 and S2) to 0.6-0.7 mu Gal (K1). The difference in amplitude among the three models is less than 0.14 mu Gal (M2), and the difference in phase is generally less than 10 degrees. In calculating the residual tide vectors using the ocean models, the TPXO6.2 model generally gave the smallest residual amplitudes. Our result for the K1 wave was anomalously large (1.36 +/- 0.25 mu Gal), while that for the M2 wave was sufficiently small (0.37 +/- 0.17 mu Gal). The associated uncertainty is half that reported in previous studies. It is interesting that the residual K1 tide is approximately 90 degrees phase-leaded, while the M2 tide is approximately 180 degrees phase-leaded (delayed). Importantly, a similar reanalysis of data collected at Asuka Station (71.5 degrees S, 24.1 degrees E) gave residual tides within 0.2-0.3 mu Gal for all major diurnal and semidiurnal waves, including the K1 wave. Therefore, the anomalous K1 residual tide observed at Vostok Station must be linked to the existence of the subglacial lake and the nature of solid-ice-water dynamics in the region. (C) 2009 Elsevier B.V. and NIPR. All rights reserved.