Functional gain (FG) and insertion gain (IG) are currently used as real ear gain measurement of hearing aids. The gain differences observed with hearing aid modifications made in order to achieve the desired prescriptive target must be greater than the variability associated with repetition of measurement in order to be really useful. Limited research is available on FG and IG intratester variability. In order to evaluate single variability factors, some studies use experimental designs not common in clinical work-up while others use equipment mot available commercially. The present study evaluates intratester variability of FG and IG (Madsen IGO 1500 equipment) in a typical clinical configuration in 42 users of behind-ear hearing aids. Mean standard deviation of FG test-retest differences was 5.19 dB. Range of variability was expressed in term of centiles. 5-- and 95-- centiles were -9.8 and 8.7 dB respectively. Mean s.d. of IG test-retest differences was 3.18 dB; 5-- and 95-- centiles were -5.2 and 6.05 dB respectively. Major variability was found at higher frequencies. High variability at lower frequencies was also found probably due to sealing problems of ear molds in the ear canal. Accord between FG and IG was also examined. A good mean correspondence in mid frequency range was found (FG-IG mean difference less than 5 dB) with large inter-subject differences between the two measurements (s.d. 11 dB).