The Theory of Visual Attention (TVA) is widely employed in both fundamental and clinical research to model attention processes and assess individual differences in attentional capacity. Given its importance, the availability of reliable and user-friendly software packages is essential for advancing TVA-based research. Among the most widely used tools for fitting TVA parameters to experimental partial and whole report data are WinTVA (Kyllingsbæk, 2005) and LibTVA (Dyrholm et al., 2011). Both are likelihood-based, platform-dependent, may require third-party software licenses, and have yet to incorporate Bayesian and/or hierarchical parameter inference. We introduce RStanTVA, a novel implementation of TVA for partial and whole report, programmed in Stan and R. This software addresses the limitations of its predecessors. Additionally, parameters can now be specified using the well-established multilevel formula syntax in R. The package enhances the flexibility and accessibility of TVA modeling by streamlining model generation, fitting, and analysis within a Bayesian framework. We replicate a range of previously published experimental findings and demonstrate the advantages of Bayesian parameter inference for TVA models, such as Bayesian hierarchical parameter inference for partial and whole report. We also discuss how a Bayesian version of TVA can drive future advancements in attention research.