Ultrasound-mediated transcranial images of the brain often suffer from acoustic distortions produced by the skull bone. In high-resolution optoacoustic microscopy, the skull-induced acoustic aberrations are known to impair image resolution and contrast, further skewing the location and intensity of the different absorbing structures. We present a virtual craniotomy deconvolution algorithm based on an ultrasound wave propagation model that corrects for the skull-induced distortions in optically-resolved optoacoustic transcranial microscopy data. The method takes advantage of the geometrical and spectral information of a pulse-echo ultrasound image of the skull simultaneously acquired by our multimodal imaging system. Transcranial mouse brain imaging experiments confirmed the ability to accurately account for the signal amplitude decay, temporal delay and pulse broadening introduced by the rodent's skull. Our study is the first to demonstrate skull-corrected transcranial optoacoustic imaging in vivo.