Scientific research in swimming over the past 10 to 15 years has been oriented toward multiple aspects that relate to applied and basic physiology, metabolism, biochemistry, and endocrinology. This review considers recent findings on: 1) specific physical characteristics of swimmers; 2) the energetics of swimming; 3) the evaluation of aerobic fitness in swimming; and 4) some metabolic and hormonal aspects related to swimmers. Firstly, the age of finalists in Olympic swimming is not much different from that of the participants from other sports. They are taller and heavier than a reference population of the same age. The height bias in swimming may be the reason for lack of success from some Asian and African countries. Experimental data point toward greater leanness, particularly in female swimmers, than was seen 10 years ago. Overall, female swimmers present a range of 14 to 19% body fat whereas males are much lower (5 to 10%). Secondly, the relationship between O2 uptake and crawl swimming velocity (at training and competitive speeds) is thought to be linear. The energy cost varies between strokes with a dichotomy between the 2 symmetrical and the 2 asymmetrical strokes. Energy expenditure in swimming is represented by the sum of the cost of translational motion (drag) and maintenance of horizontal motion (gravity). The cost of the latter decreases as speed increases. Examination of the question of size-associated effects on the cost of swimming using Huxley's allometric equation (Y = axb) shows an almost direct relationship with passive drag. Expressing energy cost in litres of O2/m/kg is proposed as a better index of technical swimming ability than the traditional expression of VO2/distance in L/km. Thirdly, maximal direct conventional techniques used to evaluate maximal oxygen consumption (VO2 max) in swimming include free swimming, tethered swimming, and flume swimming. Despite the individual peculiarities of each method, with similar experimental conditions similar results for VO2 max will be found. Free swimming (unimpeded) using the backward extrapolation method will, however, lead to reliable and valid results obtained in a condition that is closer to the competitive situation than with a direct test. A maximal indirect field-test has been recently made available. This test can predict VO2 max with an acceptable accuracy (r = 0.877), and provides a mean to evaluate the functional maximal aerobic power in swimming which corresponds to the maximal aerobic swimming velocity.(ABSTRACT TRUNCATED AT 400 WORDS)