The aim of this study was to predict indoor rowing performance in 12 competitive female rowers (age 21.3 +/- 3.6 years, height 1.68 +/- 0.54 m, body mass 67.1 +/- 11.7 kg; mean +/- s) using a 30 s rowing sprint, maximal oxygen uptake and the blood lactate response to submaximal rowing. Blood lactate and oxygen uptake ((V) over dotO(2)) were measured during a discontinuous graded exercise test on a Concept II rowing ergometer incremented by 25 W for each 2 min stage; the highest (V) over dot O-2 measured during the test was recorded as (V) over dot O (2max) (mean = 3.18 +/- 0.35 l . min(-1)). Peak power (380 +/- 63.2 W) and mean power (368 +/- 60.0 W) were determined using a modified Wingate test protocol on the Concept II rowing ergometer. Rowing performance was based on the results of the 2000 m indoor rowing championship in 1997 (466.8 +/- 12.3 s). Laboratory testing was performed within 3 weeks of the rowing championship. Submitting mean power (Power), the highest and lowest five consecutive sprint power outputs (Maximal and Minimal), percent fatigue in the sprint test (Fatigue), (V) over dot O-2max (l . min(-1)), (V) over dot O-2max (ml . kg(-1) . min(-1)), (V) over dot O-2 at the lactate threshold, power at the lactate threshold (W), maximal lactate concentration, lactate threshold (percent (V) over dot O-2max) and (V) over dot E-max (l . min(-1)) to a stepwise multiple regression analysis produced the following model to predict 2000 m rowing performance: Time(2000) = - 0.163 (Power) 14.213 .( (V) over dot O-2max l . min(-1)) + 0.738 . (Fatigue) + 567.259 (R-2 = 0.96, standard error = 2.89). These results indicate that, in the women studied, 75.7% of the variation in 2000 m indoor rowing performance time was predicted by peak power in a rowing Wingate test, while (V) over dotO(2max) and fatigue during the Wingate test explained an additional 12.1% and 8.2% of the variance, respectively.