The claim that Public Service Motivation (PSM) is an antecedent of prosocial behaviour has often been empirically tested and supported. However, closer inspection of this literature reveals large disparities in relating the two constructs. One reason that could explain such differences is that the relationship between PSM and prosocial behaviours has been primarily tested using self-reported cross-sectional, single-rater and same-survey data. While all of these are widely used methodological approaches in social sciences, they are also susceptible to potential biases. We conduct two comparative studies to re-examine this relationship. Study 1 utilizes self-reported cross-sectional, single-rater and same-survey data linking PSM and prosocial behaviour, revealing a positive relationship with PSM's Compassion dimension. Study 2 involves observing actual prosocial behaviour in a real-life setting. Then, the correlation between PSM and prosocial behaviour disappears. We conclude by discussing the possible reasons that could lead to the differences found across the two studies.