The purpose of this investigation was to compare the criterion validity and test-retest reliability of common methods for quantifying training load. Ten (5 men and 5 women) recreational athletes (mean ± SD, VO2max: 37.0 ± 4.3 ml·kg-1·min-1; age: 23.8 ± 8.4 years) completed 18 randomly assigned steady state (SS) and interval (INT) training sessions during a 6-week period. Steady-state sessions were 18 minutes in duration and were performed at 35, 50, and 65% of maximum work capacity (Wmax). Interval sessions were performed at 50, 60, and 70% of Wmax with a work to rest ratio of 1:1 and matched for total work with the 50% SS session. Oxygen consumption (VO2) and heart rate (HR) were measured throughout all sessions, whereas blood lactate concentration and rating of perceived exertion (RPE) measures were taken every 6 minutes during sessions. Session-RPE (sRPE) was collected after each exercise bout. All individual correlations between VO2 and external work (r = 0.88-0.97), HR (r = 0.65-0.90), and RPE-based methods (r = 0.55-0.89) were statistically significant. External work correlated best with the total V[Combining Dot Above]O2 and was significantly different from RPE-based methods. A poor level of test-retest reliability was shown for Banister's TRIMP (15.6% coefficient of variation [CV]), Lucia's TRIMP (10.7% CV), and sRPE (28.1% CV). Good reliability was shown for HR (3.9% CV) and a moderate level for RPE 6-20 (8.5% CV) as a measure of exercise intensity. These results suggest external work to be the most valid and reliable method for quantifying training load. Poor levels of reliability were reported for each of the HR-based TRIMP methods and RPE-based methods.