Background: Acupuncture textbooks, schools, practitioners and clinical researchers designing randomized controlled trials on acupuncture all assume that acupuncture points are small and must be located precisely.

Method: Seventy-one medical doctors with ≥200 h acupuncture training and ≥2 years of clinical experience independently identified 23 commonly used acupuncture points on a male volunteer, using sticky transparent films with an X/Y grid placed asymmetrically around acupuncture points.

Results: For each acupuncture point, the field covering 95% (68%) of all point locations varied from 2.7 (0.7) cm(2) for PC-6 up to 41.4 (10.2) cm(2) for ST-38. Commonly-used acupuncture points showed unexpectedly large variance in location: 95% (or 68%) areas were SP-6: 12.2 cm(2) (3.0 cm(2) ), ST-36: 20.7 cm(2) (5.1 cm(2) ), LI-15: 18.7 cm(2) (4.6 cm(2) ), BL-23: 22.4 cm(2) (5.6 cm(2) ) and BL-54: 22.5 cm(2) (5.6 cm(2) ). Points close to anatomical landmarks (forearm, ankle, poplitea; BL-60, BL-40, TW-5, PC-6) were located with less variance. Precision of point location was independent of length of acupuncture experience, kind of training or medical specialty.

Conclusions: In respect to the high degree of variation in the localization of acupuncture points, we suggest that the term 'acupuncture field' is more appropriate than 'acupuncture points' to describe the clinical reality; for the design of sham-controlled acupuncture trials, we recommend a minimum distance of 6 cm between verum and sham points on face, hands and feet, and up to 12 cm for all other parts of the body.