This Australian study found that voluntary error reporting and systematic, contemporaneous error detection by trained clinicians each found a different subset of errors. However, neither method alone accurately identified all errors during the study period.