OBJECTIVES: There is ambiguity concerning the walk tests available for functional assessment of coronary patients, particularly for the walking speed. This study explores the psychometric properties of two walking tests, based on fixed-distance tests, at comfortable and fast velocity, in stabilized patients at the end of a cardiac rehabilitation program. METHODS: At a three-day interval 58 coronary patients (mean age of 64.85+/-6.03 years, 50 men) performed three walk tests, the first two at a comfortable speed in a random order (6-minute walk test - 6MWT - and 400-metre comfortable walk test - 400mCWT) and the third at a brisk speed (200-metre fast walk test - 200mFWT). A modified Bruce treadmill test was associated at the end of the second phase. Monitored main parameters were: heart rate, walking velocity, VO2. RESULTS: Tolerance to the 3 tests was satisfactory. The reliability of the main parameters was good (intraclass correlation coefficient>0.8). The VO2 concerning 6MWT and 400mCWT were not significantly different (P=0.33) and were lower to the first ventilatory threshold determined by the stress test (P<0.001): 16.2+/-3.0 vs. 16.5+/-2.6 vs. 20.7+/-5.1mL.min(-1).kg(-1) respectively. The VO2 of the 200mFWT (20.2+/-3.7) was not different from the first ventilatory threshold. CONCLUSIONS: 400mCWT and 200mFWT are feasible, well-tolerated and reliable. They explore two levels of effort intensity (lower and not different to the first ventilatory threshold respectively). 400mCWT is a possible alternative to 6MWT. Associated with 200mFWT it should allow a better measurement of physical capacities and better customization of exercise training.