This study uses a new criminal case data set to partially replicate Kalven and Zeisel's classic study of judge-jury agreement. The data show essentially the same rate of judge-jury agreement as did Kalven and Zeisel for cases tried almost 50 years ago. This study also explores judge-jury agreement as a function of evidentiary strength (as reported by both judges and juries), evidentiary complexity (as reported by both judges and juries), legal complexity (as reported by judges), and locale. Regardless of which adjudicator's view of evidentiary strength is used, judges tend to convict more than juries in cases of “middle” evidentiary strength. Judges tend to acquit more than juries in cases in which judges regard the evidence favoring the prosecution as weak. Judges tend to convict more than juries in cases in which judges regard the evidence favoring the prosecution as strong. Rates of adjudicator agreement are thus partly a function of which adjudicator's view of evidentiary strength is used, a result not available to Kalven and Zeisel, who were limited to judges’ views of the evidence. We find little evidence that evidentiary complexity or legal complexity help explain rates of judge-jury disagreement. Rather, the data support the view that judges have a lower conviction threshold than juries. Local variation exists among the sites studied. The influences of juror race, sex, and education are also considered.