This paper investigates the causal effects of parental incarceration on child outcomes using linked administrative data from Wisconsin spanning court records, educational performance, and criminal system contact. Exploiting quasi-random assignment of criminal defendants to judges with varying tendencies to incarcerate, we construct an instrumental variable for incarceration that is plausibly exogenous to unobserved family characteristics, but highly predictive of incarceration on an individual case. Our analytic sample includes over children linked to a parent criminal case between 2002 and 2020. We examine a range of child outcomes, including high school completion, standardized test proficiency benchmarks, chronic absenteeism, school discipline, and criminal justice involvement. Across most domains, we find limited evidence of large causal effects. Instrumental variables estimates are generally imprecise, but allow us to rule out large negative effects on academic performance. For criminal outcomes, we find suggestive evidence that parental incarceration may reduce future system involvement among children, though estimates vary by parent gender. These findings contrast with the unconditional correlation between parental incarceration and child outcomes, which reflects unobserved selection.