The present paper investigates the acquisition of L3 Polish by L1 Italian university students of L2 Russian. The participants had never studied the L3 prior to the experiment, but took a meta-linguistically explicit course in Slavic Linguistics focussing on Polish/Russian contrastive grammar. The main research question is whether or not the similarity of the target structure in the two languages allows learners to transfer the L2 processing mechanisms to the L3. The conflicting predictions of the FT/FA hypothesis and of the DMTH are compared. The results show that alongside typical developmental errors, clear instances of L2 transfer can be identified. Although transfer does not always produce target-like results, it arguably constitutes evidence of the processing of a target structure which is not normally accessible in the early stages of acquisition.