<![CDATA[Abstract This study investigates the relationships between knowledge, attitude, perception, peer influence, participation in safety-riding training, and safety-riding behavior among students of an Automotive Engineering Education department at a university in Yogyakarta, cohort 2025. A quantitative correlational approach with a cross-sectional design was used. The entire population of 98 students was included through saturated sampling. Data were collected using closed-ended questionnaires that had been tested for validity and reliability. Univariate analysis was used to describe each variable, while bivariate relationships were examined using the Chi-Square test and Fisher’s Exact Test at a significance level of 0.05. The results showed significant relationships between knowledge and safety-riding behavior （p = 0.025）， attitude and safety-riding behavior （p = 0.030）， peer influence and safety-riding behavior （p = 0.021）， and participation in safety-riding training and safety-riding behavior (Fisher’s Exact p = 0.031）。 Perception was not significantly related to safety-riding behavior （p = 0.071）。 Descriptively,55.1%of respondents were categorized as having unsafe riding behavior, although 93.9% had previously participated in safety-riding training. These findings indicate that participation in training alone does not necessarily ensure consistent safe-riding behavior. Strengthening road-safety education should therefore combine knowledge development, attitude formation, supportive peer norms, and better-designed training programs. The findings are limited to one student cohort and rely on self-reported behavior, so broader studies using observational measures are recommended. Abstrak Penelitian ini bertujuan menganalisis hubungan pengetahuan, sikap, persepsi, peran teman sebaya, keikutsertaan pelatihan safety riding, dan perilaku safety riding pada mahasiswa Departemen Pendidikan Teknik Otomotif pada salah satu perguruan tinggi di Yogyakarta angkatan 2025. Penelitian menggunakan pendekatan kuantitatif korelasional dengan desain potong lintang. Seluruh populasi sebanyak 98 mahasiswa dijadikan sampel melalui teknik sampling jenuh. Data dikumpulkan menggunakan kuesioner tertutup yang telah melalui uji validitas dan reliabilitas. Analisis univariat digunakan untuk menggambarkan setiap variabel, sedangkan hubungan antarvariabel dianalisis menggunakan uji Chi-Square dan Fisher’s Exact Test pada taraf signifikansi 0,05. Hasil penelitian menunjukkan hubungan yang signifikan antara pengetahuan dan perilaku safety riding （p = 0,025）， sikap dan perilaku safety riding （p = 0,030）， peran teman sebaya dan perilaku safety riding （p = 0,021）， serta keikutsertaan pelatihan dan perilaku safety riding (Fisher’s Exact p = 0,031）。 Persepsi tidak memiliki hubungan yang signifikan dengan perilaku safety riding （p = 0,071）。 Secara deskriptif,55,1% responden berada pada kategori perilaku tidak aman, meskipun 93,9% pernah mengikuti pelatihan safety riding. Temuan ini menunjukkan bahwa keikutsertaan pelatihan saja belum menjamin penerapan perilaku berkendara aman secara konsisten. Program keselamatan berkendara di perguruan tinggi perlu memadukan penguatan pengetahuan, pembentukan sikap, dukungan norma teman sebaya, dan peningkatan mutu pelatihan. Hasil penelitian terbatas pada satu angkatan dan menggunakan data perilaku berbasis laporan diri, sehingga penelitian lanjutan dengan cakupan sampel lebih luas dan observasi perilaku disarankan.]]>
