本研究旨在探討機車安全教育介入對大學生交通事故傷害自覺易感性、嚴重性、安全騎機車行為自覺利益與障礙、安全騎機車之自我效能與行為意向之成效。研究採準設計，以中部某科技大學學生為樣本，共計228人。其中實驗組有122人，對照組有107人，實驗組進行課室3週6小時的「機車安全教育計畫」，對照組則未提供任何教育介入。於介入前，兩組皆施予前測，以作為學習評量之基礎。教育介入完成後立即施予後測，以了解該介入計畫之意即或效果，並於六個星期後進行後後測，以了解延宕效果。資料分析採用次數分配、平均值、卡方檢定、獨立樣本t檢定與單因子共變分析等方法來進行各項假設之考驗。研究所獲致之結論如下：立即效果方面，兩組控制在前測相同的情況下，實驗組在事故傷害自覺易感性與嚴重性、安全騎車行為自覺障礙與自我效能方面明顯優於對照組，即機車安全教育在改善交通事故傷害自覺易感性與嚴重性、安全騎車行為自覺障礙與自我效能方面具有立即的效果，但六大依變項之延宕效果則不顯著。另外參與實驗介入的學生九成以上為教學活動有實際幫助並且未來值得推廣。

The purpose of this study was to investigate the effects of motorcycle safety education program on improvement of perceived sensitivity and severity of traffic injury, perceived benefits and barriers of safety riding behavior, self-efficacy and intention of safety riding behavior. A quasi-experimental design was utilized in university students in Tai-Chung. Two hundred and twenty-eight subjects were assigned to experimental and control groups. The experimental groups, which consisted of 121 subjects, participated in a six hours program of motorcycle safety educational program. The control group, which consisted of 107 subjects, had no such program. Before the program, a pretest was conducted to establish the baseline data to compare the tow groups. After this program, posttest was conducted to understand the immediate effect. After six weeks, a post posttest was conducted to understand the delay effect. Data was analyzed by frequency, mean, chi-square test, t-test, analysis of covariance (ANCOVA). The findings and conclusions of this study were listed as follow: there were immediate effects on improvement perceived sensitivity, severity, barriers, and self-efficacy under control the pretest, but no delay effects in all dependent variables. Furthermore, over ninety percent students who participated in this program thought it would help them and worth to be used in the future.