Variability in time series of ice conditions in the Baltic Sea is examined within the context of atmospheric circulation represented by the North Atlantic Oscillation (NAO) and Arctic Oscillation (AO) winter indices using the wavelet approach. We develop methods of assessing statistical significance and confidence intervals of cross-wavelet phase and wavelet coherence. Cross-wavelet power for the time series indicates that the times of largest variance in ice conditions are in excellent agreement with significant power in the AO at 2.2--3.5, 5.7--7.8, and 12--20 year periods, similar patterns are also seen with the Southern Oscillation Index (SOI) and Ni¿o3 sea surface temperature (Ni¿o3) series. Wavelet coherence shows in-phase linkages between the 2.2--7.8 and 12--20 year period signals in both tropical and Arctic atmospheric circulation and also with ice conditions in the Baltic Sea. These results are consistent with GCM simulations showing dynamical connections between high-latitude surface conditions, tropical sea surface temperatures mediated by tropical wave propagation, the wintertime polar vortex, and the AO and with models of sea ice and oceanic feedbacks at decadal periods.