The ability of 550 consecutive patients attending a tuberculosis clinic to grade their own Heaf skin test reaction, using an illustrated card designed for such self-reporting, was assessed. Individuals were asked to grade their own Heaf skin reaction and at the same attendance the reaction was formally graded by a respiratory health visitor. There was overall excellent agreement with only 58 (10.5%) disparate gradings. We conclude that self-reporting of Heaf skin tests using an illustrated card is simple, usually accurate and may preclude an otherwise unnecessary early reattendance. This study also validates an often used system of self-reporting.