INTRODUCTION. Nowadays, accessing the European labour market opportunities implies having a high command of English. For this reason, most Spanish universities offer their Bachelor Degrees/Undergraduate Degrees in English. Learning at university, especially in distance learning, strongly depends on the student's own comprehension monitoring when reading instructional materials, usually expository texts. The present work compares comprehension monitoring (CM) in Spanish and in English of Spanish university students with a high level of English proficiency. A replication study is developed to increase the validity of the interesting results obtained on English-Spanish differences in CM at micro and macro-structural levels. METHOD. Two related empirical studies were carried out: the second one aimed at replicating the first one in a different academic context. In Study 1, thirty-three post-graduate students of a Master's Degree in Teaching Training for Compulsory Secondary Education (ESO), Upper Secondary Education (Bachillerato), Vocational Training and Language Teaching (Specialising in English Language) participated. They read three texts in English and three in Spanish in order to judge their comprehensibility. Following the error detection paradigm micro and macro-structural inconsistencies were embedded in the texts. RESULTS. Results showed that students' CM was better in Spanish than in English as expected but, in addition, an intriguing interaction effect CM-level X Language was found: in English, students' Micro-structural CM was more effective than their Macro-structural CM, whereas when reading in Spanish, the differences vanished. Study 2 was conducted with twenty-six students of a Master's Degree in Applied Linguistics. Results replicated the interaction effect found in Study1. DISCUSSION. Although students had an advanced English level, differences between the two languages, especially at macro-structural level, still remained. This suggests that English teaching has to be improved, at least in order to guarantee suitable comprehension of long texts. Results also suggest possible processing L1-FL differences. Although some hypotheses are offered to explain these differences, these should be further contrasted in future experiments.