This exploratory research paper attempted to determine the nature, causes, and instructional relevance of composition errors in a corpus of 35 compositions written by second year Engineering students of SLSU, A.Y. 2012-2013. Analyzed in reference to the Error Analysis Framework of SP Corder and Jack Richards, results revealed that seven kinds of errors were commonly found in the corpus. The seven frequently occurring errors include errors in tense sequence, word substitution, embedding, preposition substitution, spelling, and article insertion and deletion. These errors are results of over-generalization, false concept about the rule, and ignorance to rule restrictions, and are representations of the learners’ idiosyncratic dialect or interlanguage. Hence, correction of errors in tense sequence, word substitution, and embedding must be given emphasis in class discussions. Errors in preposition substitution, spelling, and article insertion and deletion may be corrected when pedagogical focus calls for discussion of such concepts.