This study was undertaken to invesigate the application of submaximal field tests in assessment of performance. A submaximal field exercise test consisting of 2 bouts of pacing for 1600 m, with 5 min rest or walking between runs was performed on Standardbred racehorses at 2 training centres (A and B). Performance indices were compiled for each horse: number of race starts, number of race wins, number of race placings (1, 2 or 3), and lifetime earnings. Regression analysis was conducted to describe the relationship between plasma lactate concentrations and speed for tests one, 2 and pooled results. Using the regression equation, observed (measured) minus expected (predicted) (O-E) lactate concentrations for tests were calculated and plotted against performance indices to determine their relationship. For stables A and B the association between lactate and velocity for the 2 runs was best described by the exponential equations: y = 0.0014(e07306x) (r2 = 0.54, P<0.001) and y = 0.002(e06812x) (r2 = 0.42, P<0.001), respectively, where y is lactate concentration (mmol/l) and x is speed (m/s). For the pooled results for heat 1 and heat 2 for stables A and B the relationship between lactate and velocity was best described by an exponential equation y = 0.0016(e07117x) (r2 = 0.47, P<0.001). This study found no relationship between (O-E) and performance indices (number of race wins, number of race placings, lifetime earnings and average $/start) at stables A or B for test run 1, 2 or pooled lactates. At stable A, there was a significant association between V(La4) and log lifetime earnings (r = 0.51, P = 0.053) and log average $/start (r = 0.54, P = 0.036). There were no significant correlations at stable B. In conclusion a 2-step determination method of V(La4) is a possible means for studying limits to performance in pacing Standardbred racehorses.