Reproduction does are commonly housed separately in cages, which have been suggested to restrict intrinsically motivated behaviours (e.g. locomotion and social contact). Using group pens instead of single-doe cages decreases spatial and physical restrictions, but continuous group housing of reproduction does can lead to infertility and high litter mortality. Therefore, we evaluated if semi-group housing (4 does housed together from 18 days after kindling until 3 days prior to the next kindling) would lead to an increase in time spent on behaviours thought to be restricted by single-doe cages. Sixteen does housed in single-doe cages (0.4 m2 + 0.1 m2 platform) and 32 does in semi-group pens (2 m2 + 0.6 m2 platform/ 4 does) were observed using continuous sampling during six 30-minute timeslots (immediately after grouping of the semi-group does, 12 hours later, and at midday and midnight 4 and 12 days later). Statistical comparisons were made using Kruskal-Wallis tests and results are presented as medians (interquartile range). In the 30 minutes immediately after grouping, semi-group does spent more time on locomotion than does in single-doe cages (4 (4-5) vs. 1 (1-1)%, P<0.01), as well as on social sniffing/grooming (1 (1-2) vs. 0 (0-0)%, P<0.01). Although such behaviours still differed significantly between semi-group and single-doe housing 4 and 12 days after grouping, the differences were much smaller (e.g. midnight 12 days post-grouping locomotion: 0.8 (0.6-1.4) vs. 0.2 (0.2-0.3)%, P<0.05; social sniffing/grooming: 0.4 (0.4-0.6) vs. 0.0 (0.0-0.0)%, P<0.01). Attacking/chasing followed a similar pattern (after grouping: 5 (4-8)% in semi-group housing vs. 0 (0-0)% in single-doe housing, P<0.01; midnight 12 days post-grouping: 0.01 (0.00-0.03) vs. 0.00 (0.00-0.00)%, P<0.10). Immediately after grouping, semi-group does spent less time in bodily contact with adult conspecifics than does from single-doe housing (who could only lie together with the wire cage wall between them, 2 (1-3) vs. 12 (11-15)%, P<0.01). This may have been due to the unfamiliarity between the grouped does. However, even 12 days after mixing the time spent in bodily contact was not significantly greater in the semi-group system than in single-doe housing (P>0.10). The limited occurrence of locomotion and non-agonistic social contact in semi-group housing suggests that either the does’ motivation for such behaviour is limited, or that the specific semi-group system we tested still imposes physical or social restrictions on such behaviour.