The aims of this study were to examine the cardiovascular response to recreational 5-a-side indoor-soccer (5v5) matches (5v5 study, 5v5S, n=15) and to assess the validity of using heart rate (HR) to estimate oxygen uptake (VO(2)) demands during actual game-play (validity study, VS, n=16) in young subjects (age 16.8+/-1.5 years). Game responses during 5v5S were assessed during 30 min matches using short-range telemetry heart-rate monitors. In VS games (12 min), VO(2) and HR were monitored with a portable gas analyser (K4b(2), COSMED, Rome, Italy). Individual HR-VO(2) relationships were determined from a laboratory treadmill run to exhaustion (VS) and a multistage shuttle running fitness test (5v5S) using K4b(2). Results showed that 5v5 elicits 83.5+/-5.4 and 75.3+/-11.2% of HR(peak) and VO(2peak), respectively. Ninety-one percent of the playing time (30 min) was spent at HR higher than 70% of HR(peak). In VS match, gas analyses revealed that only 71% of HR variance was explained by VO(2) variations. However, playing at approximately 70% of HR(peak) elicited 51.6+/-11.2% of VO(2peak). Group actual versus predicted VO(2) values demonstrated no significant differences (p>0.05), however, large confidence limits were observed (+6.20 and -10.53 mlkg(-1)min(-1)). These results show that HR and VO(2) responses to recreational 5v5 soccer in young athletes are similar to the exercise intensities recommended by ACSM for promoting cardiovascular health and suggest that HR is valid to prescribe and monitor aerobic intermittent exercise. These results also show that HR measures are acceptable for estimating VO(2) during intermittent exercise when assessing large groups, but show that large estimation errors can occur at the individual level.