The objective of this study was to evaluate the viability of using a single test in which cardiorespiratory variables are measured, to establish training guidelines in running and/or cycling training activities. Six triathletes (two females and four males), six runners (two females and four males) and six males cyclists, all with 5.5 years of serious training and still involved in racing, were tested on a treadmill and cycle ergometer. Cardiorespiratory variables [e.g., heart rate (HR), minute ventilation, carbon dioxide output ((V) over dot CO2)] were calculated relative to fixed percentages of maximal oxygen uptake ((V) over dot O-2max, from 50 to 100%). The entire group of subjects had significantly (P < 0.05) higher values of (V) over dot O-2max on the treadmill compared with the cycle ergometer [mean (SEM) 4.7 (0.8) and 4.4 (0.9) l . min(-1), respectively], and differences between tests averaged 10.5% for runners, 6.1% for triathletes and 2.8% for cyclists. A three-way analysis of variance using a 3 x 2 x 6 design (groups x tests x intensities) demonstrated that all factors yielded highly significant F-ratios (P < 0.05) for all variables between tests, even though differences in HR were only 4 beats . min(-1). When HR was plotted against a fixed percentage of (V) over dot O-2max, a high correlation was found between tests. These results demonstrate that for triathletes, cyclists and runners, the relationship between HR and percentage of (V) over dot O-2max, obtained in either a treadmill or a cycle ergometer test, may be used independently of absolute (V) over dot O-2max to obtain reference HR values that can be used to monitor their running and/or cycling training bouts.