Dairy herds in Australia calve in spring of each year, however extended lactation beyond 300 days is practiced by some dairy farmers. This study examined the impact of extended lactation on cow activity, using remote data sensors (Icetag) fitted to individual cows for 24hr. Observations were taken on 25 'Extended Lactation' cows (EL) and 25 'Control' cows (C) in autumn and the subsequent spring. In autumn the main difference between the two treatments was that the C cows but not the EL cows were in calf, while in spring the C cows were in peak lactation after having recently calved while the EL cows were producing considerably less milk, and both were back in calf. After appropriate transformation, data were analysed using REML mixed models. Results are presented as predicted medians and predicted inter-quartile ranges (IQR) on the original scale. There were no differences between the medians of the two treatments (P>0.05) in either autumn or spring between C and EL cows in number of steps/24hr (4510 vs 4580 and 4500 vs 4860 resp), %active/24hr (7.0% vs 7.5% and 7.7% vs 7.6%), %standing/24hr (53.6% vs 53.9% and 50.2% and 49.4%) and %lying/24hr (39.4% vs 37.4% and 42.1% vs 42.9%). However, in autumn there where large differences in variation in behaviour, when the IQR of C and EL cows of 'step number' was 910 vs 3010, 'active' was 0.5 vs 4.1, 'standing' was 4 vs 14 and 'lying' was 3 vs 18 (all P<0.002). These results suggest that the status of milk production and pregnancy did not affect activity and lying behaviour. However, EL cows were not pregnant in autumn and may show oestrous behaviour. This may explain the large variation of behaviour between EL cows in autumn, and therefore needs to be taken into account when analysing cow activity data.