Online teaching accelerated during COVID-19-associated lockdowns. At that time, it was assumed that university students wanted to revert to in-person sessions at the earliest opportunity. However, when in-person sessions were re-introduced, student attendance was not as high as expected. Therefore, we examined students' preferences of learning modes. Students (n = 968) from different UK universities, degree cohorts, study levels and biological sexes were given four learning-mode options: Face-to-face sessions for lectures and tutorials (in-person), Live online sessions for lectures and tutorials (Online-synchronous), Pre-recorded lectures and live online tutorials (Online-mixed-asynchronous-synchronous), and Pre-recorded lectures and face-to-face tutorials (Blended: in-person and online-asynchronous). Students ranked these options as per their preference via an online anonymous survey. Data were analysed using IBM SPSS Statistics 28. Results showed that the most frequently selected 1st and last choices were In-person and Online-synchronous modes, respectively. For the majority, above choices were the same across study levels and biological sex, but across degree cohorts, the 1st choice was either In-person or Blended. Proportion of students selecting In-person mode as their 1st choice (52.2%) was almost equal to the combined proportions of those selecting other learning modes as 1st choices (47.5%). Amongst degree cohorts, In-person mode was least preferred by Language Education students and most preferred by Bioscience and Sports & Exercise Science students. The latter cohort also preferred Online-synchronous mode more than other degree cohorts. Blended mode was preferred more by Language Education, Computer Science and Psychology students but preferred less by Sports & Exercise Science and Pharmacy students, compared to other degree cohorts. Ordinal regression revealed that Sports & Exercise Science students preferred Online-mixed-asynchronous-synchronous mode less than Language Education students. Undergraduates preferred In-person mode more and Online-mixed- asynchronous-synchronous mode less than postgraduates. Preference differences between biological sexes were insignificant. Thus, we identified students' preferences of learning modes and propose that not biological sex, but discipline and study level can predict/influence preferences.