There have been a considerable number of studies conducted on dairy calves housed in pairs, which is a promising alternative to individual housing. Most studies report positive effects of pair housing on calf welfare, but only a few compared how these calves differ in play. Therefore, we aimed to investigate play in individually versus pair-housed calves. We hypothesized that individually housed calves will perform less locomotor as well as total play due to space restriction and absence of a conspecific. Additionally, we hypothesized that individually housed calves will perform more object play as a substitute for reduced locomotor and total play. The study was carried out at the Institute of Animal Science’s experimental farm Netluky in Prague, Czech Republic, from June 2017 to August 2018. In total, 58 Holstein Friesian female calves were used. Calves were separated from their mothers within 12 hours after birth. After 1 wk of individual housing, 40 calves were assigned to pairs (PAIR) while 18 calves continued individual housing (INDI). Home-pen behavior was recorded by camera system from 3 to 8 wk of calf age. One hour per week (6 h in total/calf) was observed by a single observer using One-zero sampling method. This hour was divided to 1 min sampling intervals, i.e., we recorded if the behavior was present at any time during the 1-min interval. Play behaviors of interests were gallop, jump, buck-kick, turn (locomotor play), butting or rubbing straw, and butting fixtures (object play). Additionally, frontal pushing, body pushing and mounting were observed in PAIR calves. The sums were calculated for each play category, and total play was calculated as sum of all play behaviors. The observer achieved high intra-observer reliability in all play categories (r > 99%; calculated by Pearson correlation coefficient). All statistical analyses were performed using generalized linear models with generalized estimating equations (GEE). The models were fitted using a Poisson distribution with a log link function, except for the model testing the occurrence of object play, where a binomial distribution with a logit link was used. Data was analyzed using SAS 9.4. Housing did not affect total (DF = 1; χ2 = 0.13, p=0.71) and locomotor (DF = 1; χ2 = 0.20, p=0.66) play. However, we found that housing significantly affected occurrence of object play (DF = 1; χ2 = 4.32, p=0.038). Predicted probabilities showed that individually housed calves had a higher probability of object play (0.51 +- 0.08) compared to pair-housed calves (0.26 +- 0.07). Because total and locomotor play did not differ between INDI and PAIR calves, our results point out that play may serve as a „buffer“ against stress in some situations where animal welfare may be at stake (e.g., social isolation). This study was approved by the Institutional Animal Care and Use Committee of the Institute of Animal Science.