The objective of this study was to investigate the effect of different trunk postures during the abdominal flexor test on isometric endurance holding times and surface electromyographic (EMG) activity of contributing trunk flexor and extensor muscles in 16 healthy college-age male rowers (age, 22.0 ± 1.55 years; height, 179.1 ± 5.47 cm; weight, 68.0 ± 4.18 kg) with a minimum of 12 months rowing experience. Subjects performed 2 maximal effort abdominal flexor tests, 1 held with correct posture and the other with a slightly flexed incorrect trunk posture. Maximum holding time was assessed for each posture while EMG muscle activity was recorded for the external oblique, rectus abdominis, internal oblique, rectus femoris, and lumbar erector spinae muscles. Mean maximal holding time for the incorrectly positioned abdominal flexor test was significantly longer than for the correct position (p < 0.05). EMG activity during the first and last 20% of a time normalized comparison showed significantly lower muscle activity during the incorrect trunk posture in all muscles investigated (p < 0.05). There were no significant differences (p > 0.05) in muscle activity when comparing an equivalent time period at the actual termination of each of the 2 abdominal flexor tests. Although minor variations in trunk posture seem to be acceptable in practice of this isometric abdominal flexor test, the present data indicate that minor lumbar kyphosis deviating away from the correct abdominal flexor test posture can significantly skew test reliability, resulting in longer endurance holding times and relatively lower muscle activity of the trunk muscles.