The purpose of this trial was to determine whether cobedding of preterm twins has analgesic effects during heel lancing or not. Methods: One hundred premature twins (50 sets) born between 26 weeks' and 34 weeks' gestation undergoing heel blood sampling were randomly assigned into two groups: the cobedding group (receiving care in the same incubator) and the standard care group (receiving care in separate incubators). Pain was assessed using the premature infant pain profile score. Duration of crying was measured after heel blood sampling, and salivary cortisol was measured prior to and after heel blood sampling. Results: Infants in the standard care group cried for a longer time during heel lancing than those in the cobedding group (42.6 ± 19.8 seconds vs. 36.4 ± 21.7 seconds, p = 0.03). The mean premature infant pain profile score after heel lancing was significantly higher in the standard care group (9.8 ± 2.6 vs. 8.06 ± 2.8, p = 0.002). The mean salivary cortisol after heel lancing was also significantly higher in the standard care group (24.3 ± 7.4 nmol/L vs. 20.8 ± 7.4 nmol/L, p = 0.02). No significant adverse effects were seen with cobedding. Conclusion: Cobedding is a comforting measure for twin premature infants during heel lancing, which can be performed without any significant adverse effects.