Seven female and eight male elite junior skaters performed cycle ergometer tests at four different times during the 1987/1988 season. The tests consisted of a Wingate-type 30-s sprint test and a 2.5-min supramaximal test. The subjects were tested in February, May and September 1987 and in January 1988. Maximal oxygen consumption was measured during the 2.5-min test. With the exception of the maximal oxygen consumption of the women in May which was about 6% lower than in the other three tests, no seasonal changes in the test results could be observed--this, in spite of a distinct increase in training volume (from 10 to more than 20 h.week-1) and training intensity in the course of the season. When the test data were compared to those of elite senior skaters, it appeared that the junior skaters showed the same values for mean power output during the sprint test [14.2 (SD 0.4) W.kg-1 for the men and 12.6 (SD 0.5) W.kg-1 for the women] and maximal oxygen consumption [63.1 (SD 2.8) ml.kg-1.min-1 for the men and 55.3 (SD 3.5) ml.kg-1.min-1 for the women, respectively] as found for senior skaters. It seemed, therefore, that the effects of training in these skaters had already levelled off in the period before they participated in this investigation. In contrast to previous studies, no relationship could be shown between the test results and skating performance. This was most likely due to the homogenous character of the groups (mean standard deviations in power and oxygen consumption were only 5%).(ABSTRACT TRUNCATED AT 250 WORDS)