Coherence is an inherent element of all effective written communication and, hence, this study purports to investigate the manifestation of textual coherence types in a corpus of fifty answer sheets produced by Sudanese English majors at a large public university. The methodological underpinnings are based on Halliday’s (1994) conception of the clause as comprising an information structure and a theme—rheme organization representing the topical constituents, while at the macro- textual level, Fahnestock’s (1983) elaboration of continuative clauses was utilized. The analysis of the clausal and textual coherence was undertaken on the syntactic, lexical and discoursal errors embodied in students’ texts, accompanied by a detailed study highlighting instances of coherence breaks. It was revealed that the overall coherence of the corpus was negatively affected by the innumerable errors committed at the syntactic and lexical levels. Intra- sentential clausal coherence, for example, was compromised by the incapacity to unambiguously distinguish between given and new information, due to the subjects’ countless errors in using definite, indefinite articles and pronominalization processes. Equally, the information structure was often disrupted by faulty use of the passivization function. Regarding inter- sentential textual coherence, it was generally vitiated by such factors as students’ poor handling of conjunctions, as reflected in confusion of the basic rhetorical typologies such as cause- result and conditional- consequence, a situation aggravated by the students’ verbose writing style and their inattention to textual boundaries and macro- structure. The prevalence of the two major error categories has ultimately resulted in low interpretability, semantic ambiguity and textual opacity.