<p>Animals may acquire information about their environment by social learning. Social transmission can affect the rate of trait acquisition and performance. It is often unclear, how behaviours are acquired when social information is available. In particular, the role of social learning in the acquisition of non-intuitive tasks is currently obscure. We asked whether wild-type Norway rats, <i>Rattus norvegicus</i>, in six semi-natural outside colonies benefit from each other in the acquisition and performance of a non-intuitive task by social learning. The task involved the opening of a seesaw mechanism to obtain a food reward. We induced innovations in four of six colonies and controlled the number of trained individuals and the relatedness composition. The latency to the first successful seesaw manipulation was shorter for naïve rats living with four experienced rats than for those living with zero or two experienced rats. The intervals between successful seesaw manipulations were not affected by the number of experienced rats and the colonies’ relatedness composition. Rats in four colonies innovated seesaw manipulations by trial-and-error learning. Our data show that information about the solution of a non-intuitive food acquiring problem can be socially transmitted among wild-type Norway rats, irrespective of the colonies’ relatedness composition.</p>