Methods: A cross-sectional study was performed from January to April 2020 among students at Jahangirnagar University (JU), Bangladesh. Purposive sampling was conducted through an in-person interview using a structured questionnaire. Students from the faculties of biological sciences and non-biology background were included. The univariate ordinal regression technique was used to analyze the relationship between predictors and good knowledge about the antibiotics. A two-tailed p-value was calculated to determine statistical association.

Results: Out of 205 study participants, 92 and 113 responders were from biological science faculty and non-biology disciplines, respectively. Less than half of the students (42.4%) showed a good knowledge level (scores higher than 80%). Biology-background students possess better knowledge than non-biology students [odds ratio (OR) = 4.44, 95% confidence level (CL) (2.56, 7.70), p < 0.001]. A better attitude was noticed among all students. The self-medication rate was quite low, and more than 90% of students were found to consume antibiotics according to the physician's prescription. Lack of treatment adherence was recorded, and students admitted to stop-taking antibiotics when symptoms disappeared (48.67% biology and 36.26% non-biology). Multivariate regression analysis was unable to detect any significant association between self-medication and gender, student category or the level of knowledge about antibiotics.