Background The two main approaches in aphasia treatment are cognitive–linguistic treatment (CLT), aimed at restoring the linguistic levels affected, semantics, phonology or syntax, and communicative treatment, aimed at optimising information transfer by training compensatory strategies and use of residual language skills. The hypothesis that CLT is more effective than communicative treatment in the early stages after stroke was tested in this study.

Methods In this multicentre, randomised, parallel group trial with blinded outcome assessment, 80 patients with aphasia after stroke were included within 3 weeks post-stroke. Patients received 6 months of CLT, comprising semantic and/or phonological training, or communicative treatment for at least 2 h per week. They were assessed before treatment and at 3 and 6 months with the Amsterdam–Nijmegen Everyday Language Test (ANELT-A, primary outcome) and semantic and phonological tests (secondary outcomes). The intervention effect was evaluated by means of analysis of covariance, with adjustment for baseline scores.

Results There was no difference between the mean ANELT-A score of the CLT group (n=38) and the communicative treatment group (n=42), at 3 months (adjusted difference 1.5, 95% CI −2.6 to 5.6) or at 6 months (adjusted difference 1.6, 95% CI −2.3 to 5.6) post-stroke. On two of six specific semantic and phonological tests, the mean scores differed significantly, both in favour of CLT.

Conclusion This study does not confirm the hypothesis that patients with aphasia after stroke benefit more from CLT, aimed at activation of the underlying semantic and phonologic processes, than from general, non-specific communicative treatment (ISRCTN67723958 Current Controlled Trials).