OBJECTIVE : We compared our historical medium-term data obtained with an active semi-implanted bone conduction device and the hearing results of a new passive bone conduction hearing device to determine its predictive value for the hearing results with the semi-implanted device. METHODS : The study sample was 15 patients with an active bone conduction implant (mean follow-up 26 months). Pure tone audiometry was performed with headphones, sound field speech audiometry was conducted unaided, and free-field speech audiometry was carried out with both the active bone conduction system and the passive device switched off. RESULTS : As compared with the unaided condition, speech reception was significantly improved with both devices. Comparison of speech reception threshold at 100% of word recognition showed no difference between the active and the passive device. At lower intensity the difference in speech perception was significant in the patients with monaural fitting (group A) and was non-statistically significant in those with binaural fitting (group B); the speech reception threshold at 50% of word recognition was 26.00 dB (± 10.22) with the active implant and 30.50 dB (± 7.98) with the passive device in group A (p = 0.047) and 24.00 dB (± 5.48) and 29.00 dB (± 2.24) in group B (p = 0.052), respectively. CONCLUSIONS : The hearing outcome after active bone conduction implant was comparable to published data. Compared with the unaided condition, speech recognition was significantly improved with the passive device. The device may also provide value to predict the hearing outcome with the implanted device, especially at higher intensities. METHODS : IV.