The idea that NK cells can distinguish aberrant cells by recognizing 'absence of the expected', rather than 'presence of the unexpected' emerged more than 25 years ago. Klas Kärre recapitulates how the idea took shape, and the first five years of experimental work to test its general predictions.