Abstract: Interbed multiple suppression is a critical step affecting the accuracy of seismic data processing and interpretation. The virtual events theory offers a methodological basis for fully data-driven suppression of interbed multiples. However, the adaptive subtraction process tends to damage primary reflections when interbed multiples interfere with primaeies. The adaptive virtual events method, despite employing an iterative strategy, does not fully eliminate its reliance on adaptive subtraction. To overcome this limitation, we propose an inversion-based adaptive virtual events method that constructs an objective function to directly estimate interbed multiples through waveform inversion and thereby avoids adaptive subtraction. This method effectively suppresses interbed multiples while preserving primary energy. Experimental results demonstrate that the proposed method yields higher precision in suppressing interbed multiples and better preserves primary amplitude compared to the conventional adaptive virtual events method, thereby enhancing the data quality for subsequent seismic data processing and interpretation.