Emergent scholarship challenged the Weberian wisdom that suggests that non-meritocratic appointments are detrimental for institutional performance. In the Mexican judiciary context, this paper explores when and why we might expect positive court's performance to occur, even in the presence of high levels of nepotism. I argue that the type of family network structure shapes performance incentives: different configurations of family network density and centralisation promote accountability, monitoring and performance of their relatives. At first, I find support for the Weberian account: a) Courts with no family ties, on average, perform better than courts with family ties regardless of the family network structure. However, when looking at only courts with family ties, b) there is a statistically significant nonlinear U-shaped relationship between familial patronage and judicial performance; and c) the structure of family network produce differentiated effects on efficiency: horizontal -rather than vertical- family networks within courts produce positive effects on judicial performance. This paper contributes to the growing body of literature on positive patronage, and it introduces tools for analysing informal power structures in the judiciary through family network dynamics.