Fifteen male runners were tested on two occasions to determine the reliability of a continuous incremental level running treadmill protocol (C), with 3-min stages, for lactate threshold (LT) and fixed blood lactate concentration (FBLC) (2.0 mM, 2.5 mM, 4.0 mM) assessment. Test-retest (T-RT) reliability coefficients for velocity at LT, 2.0 mM, 2.5 mM, and 4.0 mM were r = 0.89, 0.91, 0.95, and 0.95, respectively (velocity ranged from 215.3 m/min at LT to 273.6 m/min at 4.0 mM). Mean differences in T-RT velocity values ranged from 0.7 m/min (at LT) to 6.0 m/min (at 2.5 mM; 252.3 m/min vs 258.3 m/min; NS) and the standard errors of measurement were less than +/- 10.0 m/min. Similar results were observed for VO2, with T-RT r values ranging from r = 0.82 (at LT) to r = 0.88 (at 2.0 mM) (VO2 ranged from 47.2 ml/kg.min-1 at LT to 60.9 ml/kg.min-1 at 4.0 mM; VO2 peak = 65.6 ml/kg.min-1). Mean differences in T-RT VO2 values were less than 1.4 ml/kg.min-1 (NS) and the standard errors of measurement were less than +/- 2.95 ml/kg.min-1. Interinvestigator and intrainvestigator reliability coefficients were high and ranged from r = 0.91 to r = 0.99. Validity of C was assessed in an additional 16 male runners who completed C and a criterion discontinuous (D) protocol (using a series of 10-min stages) for LT and FBLC assessment. Results indicated that C and D resulted in similar VO2 and velocity values at LT and FBLC.(ABSTRACT TRUNCATED AT 250 WORDS)