In our News Feature “Mighty mouse” (Nature 444, 814–816; 2006) we erroneously stated that 42% of the 50 knockout mouse lines studied showed new phenotypes. The correct figure is 47 lines out of the 50 (94%).