Palaeomagnetic measurements were made for 11 post-glacial basalt lavas in Iceland: three pre-2 ka lavas, seven historical lavas and one present-day lavas. Palaeodirections were easy to determine at most sites, but three of the sites seem to have suffered from some block movement at the lava front, making determination of palaeodirections difficult. Reliable site mean palaeodirections were successfully obtained at two of these sites by analysing the remanence components with low-blocking temperature, but one site was rejected due to low convergence of remanence directions, indicating a possible post-cooling block movement. A total of 45 samples (52 specimens) from nine lava flows were selected for palaeointensity experiments using the Thellier method based on the thermal stability of remanence and other rock magnetic features together with hysteresis parameters to show single-domain to pseudo-single-domain natures. 26 of the 45 samples (29 of the 52 specimens) produced successful results. Subsidiary experiments using Shaw’s method were also carried out on one specimen from each of eight lava flows, and all of these produced successful results. Combining these Thellier and Shaw results, eight site mean palaeointensities were obtained, including one 1984 AD lava, giving a specimen level success rate of 56 per cent for the experiments using the Thellier method and 100 per cent for those using Shaw’s method. Our results together with those of previous studies indicate that variations in palaeointensity and palaeodirection in Iceland are basically similar to those in NW Europe for the last 2000 yr. Variations in palaeointensity, converted to virtual dipole moment or virtual axial dipole moment, are in good agreement with both NW Europe and global models. Some differences in inclination and declination are recognized between Iceland and NW Europe, but overall variation curves are similar and good agreement with the global models is also observed.