The effects of early handling on play behaviour, faecal corticosteroid metabolite concentrations and social interactions during regrouping was studied in 40 Holstein Friesian group housed dairy calves. Calves were exposed to either positive (handler speaking quietly and using pats, n=20) or negative (pushing the animals and using harsh voices, n=20) handling for five minutes, twice a day, for five weeks (1-5 weeks of age). Fresh faecal samples were collected daily from each group for corticosteroid metabolite analysis. Play behaviour was video recorded for 30 minutes, once a week, for five weeks. There were no differences in the frequency of any play behaviour between the positive and negative treatments (mean values; positive: 2.5 events/animal/30 minutes, negative: 3.1 events/animal/30 minutes, P greater than or equal to 0.122). Play running increased by 12 seconds between weeks one and two (SED: 2.0, P=0.000) and decreased by 11 seconds between weeks four and five (SED: 2.0, P=0.022) when treatment groups were combined. The negative groups tended to have higher maximum concentrations of corticosteroid metabolites (positive: 77 ng/g, negative: 148 ng/g, SED: 18.8, P=0.064) during the first week of handling than the positives (two groups/treatment). In a follow up study conducted at nine months of age, the initial 40 animals and 20 control animals (i.e. animals that received normal on-farm handling) of the same age were assessed on their fearfulness of humans (avoidance test, and latency to approach a food container with and without a handler present), and their social behaviour during regrouping, focussing on aggression. There were no treatment differences in the latency to approach the food container (latency to approach container with and without handler, positive: 21 seconds, negative: 22 seconds, control: 23 seconds, SED: 4.0, P=0.365) and avoidance test scores (positive: 1.3, negative: 1.2, control: 0.9, SED: 0.26, P=0.528); nor were there any differences in behaviours exhibited during regrouping (eating: 11, 10, 10 No/animal/15 minutes, SED : 1.3, P=0.680, head to head duration greater than 1 second: 0.7, 0.6, 0.3 No/animal/15 minutes SED: 0.14, P=0.630, agonistic behaviours: 1.4, 1.6, 2.2 No/animal/15 minutes, SED: 0.42 , P=0.290, for positive, negative and control, respectively). Unfamiliar animals performed more head to head behaviours (duration greater than 1 second; familiar: 0.44 No/animal/15 minutes, unfamiliar: 0.94 No/animal/15 minutes, SED: 0.18, P=0.049) than familiar animals (positive, negative and control). In conclusion, early handling had no effect on play behaviour in dairy calves; however negatively handled calves tended to have higher faecal corticosteroid metabolites. Also there were no detectable differences in calves‟ fearfulness of humans or social behaviours at nine months of age; however positively and negatively handled animals showed less avoidance behaviours and performed fewer aggressions compared to animals that received minimal handling (control and unfamiliar: trend only). These results could have been influenced by the animals habituating to the negative handling or the length of time between handling application (one to five weeks of age) and social behaviour testing (nine months of age). However, further research is needed to determine how the amount of handling received at an early age influences fearfulness, play and social interactions in calves.