Objectives: The active middle ear implant, Vibrant Soundbridge (VSB), can be implanted with a variety of couplers. Hearing outcome after implantation has been investigated in both temporal bone (TB) experiments and patient studies, but the relationship between experimental and clinical data is still weak in the literature. Therefore, experimental data from TB experiments should be compared with patient data in a retrospective study, in which the floating mass transducer is used with couplers of the third generation. Actuator coupling structures included the long (LP coupler) and short (SP coupler) incus process, the stapes head (Clip coupler), and the round window membrane (RW soft coupler). Methods: In the TB experiments, the sound transmission after vibroplasty on the above-mentioned actuator coupling structures was determined in 32 specimens by means of laser Doppler vibrometry on the stapes footplate. Data of 69 patients were analyzed. The main target audiometric parameters were the postoperative aided word recognition score (WRS) in the free field at 65 dB SPL (WRS 65 dB in %), the preoperative and postoperative pure-tone average (PTA4, including the frequencies 0.5, 1, 2, and 4 kHz) of the bone conduction hearing threshold (PTA4BC), the aided postoperative air conduction hearing threshold in the free field (PTA4FF) and the direct threshold (Vibrogram) at least 6 months postoperatively. The coupling efficiency of the actuator (Vibrogram - PTA4BC) as well as the effective hearing gain (PTA4FF - PTA4BC) was compared between the couplers. Results: The analysis in the main speech range (0.5-4 kHz) indicated that in the TB experiments, the LP coupler tends to have the best coupling quality at low frequencies (500-1000 Hz). This was up to 15 dB above the worst actuator (RW soft coupler). However, the results missed the significance level (p > 0.05). In the high frequencies (2000-4000 Hz), the Clip coupler showed the best coupling quality. This was 15 dB above the worst actuator (SP coupler). However, the results missed the significance level (p > 0.05), too. The postoperative WRS at 65 dB SPL and the postoperative PTA4FF were independent of the actuator coupling structure. The PTA4BC was stable at 6 months postoperatively. For the PTA4 of the coupling efficiency, there were no significant differences between the actuator coupling structures (LP 8.9 dB ± 12.9; SP 9.5 ± 6.5 dB; Clip 5.2 ± 10.5 dB; RW 12.7 ± 11.0 dB). However, the tendential inferiority of the RW soft coupler with regard to transmission in the low-frequency range and the tendential superiority of the Clip coupler in the high-frequency range that have already been displayed experimentally could be confirmed in the clinical results. However, the clinical results missed the significance level, too (p > 0.05). Conclusions: In vivo, there are no significant differences in the postoperative outcome stratified according to coupling the target structure. The differences known from the experimental setting were repressed by individual biasing factors. However, to ensure sufficient postoperative speech intelligibility, the frequency-specific transmission behavior of the couplers should be taken into account when setting the indication for VSB implantation.