Race analysis has shown convincingly that times spent in the turning and starting phases of swimming races are strongly related to swimming performance (for example, Arellano at al, 1994; Mason and Cossor, 1999). Lyttle and Blanksby (2000) have shown that speed of the swimmer following a turn or immediately after entry in the starting dive is greater than the speed that can be achieved in underwater kicking. They found a maximal speed kicking underwater of between 1.9 and 2.2 m.s-1 in their group of experienced adult male swimmers. Kicking at higher speeds increases drag compared to that in a passive streamlined position. This has important implications. A swimmer can lose speed unnecessarily if the kick is initiated prior to the maximum speed that can be sustained by kicking. Similarly, commencing the kick too late will result in a loss of speed. Kicking too early has the added disadvantage of wasting energy. We have developed an objective method of quantifying speed and estimating appropriate times of initiation the kick. This method can be applied to advise competitive swimmers regardless of age, ability, and gender. Conclusion: The development of a mathematical model and analysis program based on fitting the glide period of a turn or dive has been found useful in assessing the timing of swimmers kicking in turns and starts. This model can be applied to determine how well swimmers can maintain their speed through streamlining as well as whether they are kicking too early or too late. From our analyses of competitive swimmers to date, it appears that kicking too early is very common. This reduces speed as well as wasting energy unnecessarily.