Several theories suggest the existence of a generalized magnitude representation system for various magnitude dimensions (e.g. A Theory of Magnitude, Walsh, 2003). Furthermore, prior studies indicate an association between magnitudes like numbers and the auditory intensity dimension loudness (e.g. Hartmann & Mast, 2017). If loudness is also represented on a generalized magnitude representation system, loudness should influence the processing of other magnitudes even if they are presented in a non-auditory modality. To investigate this assumption, we tested the influence of loudness on the processing of visually presented numbers in two experiments (online, N = 33 and a lab-based replication, N = 27). In each trial of both experiments, participants saw a single number and they had to decide whether the presented number was smaller or larger than five by pressing a left-sided or right-sided response key. Simultaneously with the presentation of the number, participants heard a tone which could be either loud or soft. We found a significant congruency effect with faster reaction times in congruent conditions (large number and loud tone or small number and soft tone) in comparison to incongruent conditions in both experiments. There was also a significant interaction between loudness level and the numerical distance between the presented number and the standard five in the online experiment but not in the lab-based experiment. We will present possible explanations for this difference and discuss the results’ implications for an assumed generalized magnitude representation of loudness.