From calving, dairy cows are typically milked for about a year, and subsequently managed to have a non-lactating or ‘dry period’ (DP) before next calving. However, the use of a DP may reduce cow welfare because typical DP management involves the cow changing groups and ration. Also, the DP results in a severe negative energy balance after calving. Shortening or omitting the DP may have beneficial effects on cow welfare through fewer changes in management before calving, and a lower milk yield after calving. Our objective was to assess the effects of no DP and a short DP (30 days) with associated management on feeding, lying, and number of steps of dairy cows in late gestation and early lactation. Feeding behaviour was recorded by computerized feeders for 122 periods (42 with a short DP and 80 with no DP) from week −6 to week 7 relative to calving. Steps and lying behaviour of 81 of these cows (28 with a short DP and 53 with no DP) were recorded with accelerometers in week −4 and in week 4 relative to calving only. Effects of DP treatment and parity on behaviour were analysed with mixed models. Before calving, cows with a short DP were fed a DP ration, and moved to a dry cow group. During this time, cows with a short DP spent more time lying (13.7 vs. 12.6 h per day; P = 0.01) and feeding (240 vs. 209 min per day; P < 0.01), and stepped less (663 vs. 1130 steps per day; P < 0.01) than cows with no DP. After calving, all cows were fed the same lactation ration and were housed in the same herd. Cows with a short DP, however, had a lower feed intake (35.7 vs 39.1 kg per day; P < 0.01), and spent less time lying (10.7 vs. 11.6 h per day; P = 0.03) after calving than cows with no DP. Milk yield was negatively correlated with daily lying time (r: −0.22; P < 0.05), but was not correlated with daily feeding time. Also, less time was spent on both lying and feeding after calving than before calving. These results indicate that lying time was not constrained by feeding time. Lying time was positively correlated with energy balance (r: 0.28; P < 0.01). Compared with a short DP with associated ration and group changes, no DP reduced lying time and increased the number of steps in late gestation, and resulted in a higher feed intake and longer lying time in early lactation.