自由車運動已在國內發展了很長的一段時間，但是一直沒有出現突破性的國際級水準成績，反觀同樣屬於東方人體型的中國大陸及日本等國的自由車選手，卻有能力在國際比賽奪取獎牌，證明自由車運動是適合東方人體型的運動。想要提高自由車運動員的成積表現，有系統的訓練是必要的，為了使訓練負荷的安排與訓練方式的選擇能夠針對相對應的運動項目特點，對各種訓練方式進行科學化的對應成效分析成為系統訓練中不可是缺的一部分。本研究特別針對男子自由車場地賽中一千公尺個人計時賽及四千公尺追逐賽的成績與各種訓練方式之間的關係，利用Microsoft Excel 97中所提供的資料分析程序逐項進行迴歸分析，獲得了彼此之間的相關係數。利用初步分析所得的結果，本研究再分別針對一千公尺個人計時賽及四千公尺追逐賽的成績，選出數項相關性較高的訓練方式為變數進行多變數線性迴歸分析，以獲得各自的線性迴歸方程式。本研究建議教練員應該針對各個選手建立其獨一無二的專項運成績與各種訓練方式之間的線性迴歸方程式，如此才能準確的預測運動員的專項成績，並且以此為依據來選擇訓練方式及安排訓練負荷，使得運動員的訓練過程達到最佳化的目的。

Until now, cycling sport in Taiwan hasn't launched the outstanding records reaching the international standard even though it has developed for a long period of time in Taiwan. Nevertheless other cyclists with the same Asian body shape, such as those from Mainland China and Japan, show their capability winning medals in the international cycling competitions; this does prove that cycling sport is a sport for Asian-body-shape cyclists. In order to better the presentation and enhance records of cyclists in Taiwan, it is necessary to implement a systematic training. Thus, to carry on the scientific correlated analysis toward various kinds of practices turns to be the indispensable part in the systematic training for the purpose of well arranging the burden resulted from cycling and choosing the best training approaches to correspond the certain sport items. This research, in regarding to the relation between the male l000-meter individual time trail race and the male 4000-meter pursuit race and all the practices related to these two races, makes a use of the information-analyzing procedure in Microsoft Excel 97 and then operates linear regression analysis in order to obtain the correlation coefficient. After gaining the result from the preliminary analysis, for the male 1000-meter individual time trial race and the male 4000-meter pursuit race respectively, the training approaches with the higher correlation efficient are chosen as variables in order to implement the multiple linear regression analysis for the purpose of getting the function of linear regression analysis of each race. This research suggests that coaches set up a unique function of linear regression analysis for the record of certain item and various practices for every cyclist. In this way, they can anticipate precisely those cyclists' records for their certain items and, according to the records, choose the best training approaches and arrange the cyclists' burden while practicing in pursuit of helping cyclists reach the best condition during the process of practicing.