This study evaluated energy cost and energy sources of a ballet exercise (grand adage) in young female dancers with different technical ability, and then related the energy sources to the subject's VO(2max) and anaerobic threshold (IAT). Twenty-five dancers (13-16 years) were divided into two different technical ability groups: low-level (n = 13) and high-level (n = 12). The overall energy requirement of dance exercise (VO(2eq)) was obtained by adding the amount of VO(2) during exercise above resting (aerobic source or VO(2ex)) to the VO(2) up to the fast component of recovery (anaerobic alactic source or VO(2al)) and to the energy equivalent of peak blood lactate accumulation (anaerobic lactic source or VO(2la) of recovery. VO(2eq) of exercise amounted to 81 +/- 10 and 94 +/- 9 ml kg(-1) in low-level and high-level groups, respectively. VO(2ex) represented the higher fraction (65 +/- 4% and 77 +/- 5%) in low-level and high-level groups, respectively, of VO(2eq )in both the groups. In the low-level group the remaining fractions were: 23 +/- 2 % for VO(2al) and 12 +/- 1% for VO(2la). In high-level group the remaining fractions were: 18 +/- 2 % for VO(2al) and 4 +/- 1% for VO(2la). Between two groups, significant differences were found in VO(2ex )(P < 0.01), VO(2al) (P < 0.01), and VO(2al )(P < 0.05). IAT was 55 and 60% of VO(2max) for low-level and high-level dancers, respectively. Low-level dancers performed more exercise above IAT than high-level. For these reasons, it should be better to define exercise intensity according to the IAT parameter and not only to VO(2max).