In this study the ice drift in two very different regional ice zones is compared. The study is based on the ice-drift observations made with drifting ice buoys in the Sea of Okhotsk during the winters of 2004 and 2005 and with the ship moored into the drifting ice floe in the Baltic Sea in March 2009. Ice drift was compared with wind observations in both sea areas. In addition, in the Sea of Okhotsk sea-level measurements and the tidal component of sea-level variations were used to study the importance of tidal forcing. During the Baltic Sea ice-drift experiment, data collected in three drifting ice stations represented three very different drifting conditions. At the first ice station (S1) the wind was mild and the icefield was very compact resulting in very slow drift. In the second ice station (S2) stronger wind caused faster drift, but also a lot of ice deformations. Fastest ice drift (22 cm s^–1) was observed in the third ice station (S3), where ice drift was very close to free drift. In S1 and S3 power spectrums of wind and ice drift correspond well over the whole frequency span, while in S2 correspondence especially in higher frequencies is clearly weaker and high rates of deformations lower the power of ice drift. In the Sea of Okhotsk ice was drifting generally much faster than in the Baltic Sea. Although all of the ice-drift power spectrums do not show very clear tidal peaks, cross wavelet and wavelet coherence analyses show the importance of tides. In the cross wavelet transforms of ice drift and sea-level variations there is significant common power in tidal bands of about 12 and 24 hours, and the common power varies with varying maximum tide amplitude.