Purpose: The aim of this study was to prove the efficacy of supervised self-training for individuals with aphasia. Linguistic and communicative performance in structured dialogues represented the main study parameters. Method: In a cross-over design for randomized matched pairs, 18 individuals with chronic aphasia were examined during 12 weeks of supervised home training. Intensive language training, assisted by an electronic learning device (B.A.Bar), was compared with nonlinguistic training. Language performance, communicative abilities, and cognitive abilities were controlled before and after each intervention and at follow-up. The language training was designed to facilitate dialogue skills as required in everyday life. Results: Robust and specific improvements in the participants' linguistic and communicative abilities were obtained using B.A.Bar dialogue training but not with nonlinguistic training. The transfer to general linguistic and communicative performance remained limited when the whole group was considered. For 30%-50% of the participants, individual analysis revealed significant improvements in spontaneous language and general communicative skills. Furthermore, individual participants demonstrated significant improvements regarding standardized aphasia assessment and proxy rating of communicative effectiveness. Conclusion: Supervised home training works. This study has proven that it is an effective tool for bolstering linguistic and communicative skills of individuals with aphasia.