Research to date has commonly suggested that meaning inferencing through concordance lines can facilitate vocabulary learning. This facilitative role, however, may be subject to mediation by the expanded contexts of the target vocabulary item in concordance and accurate meaning inferencing. Of these plausible factors, the length of the co-text of vocabulary items in concordance context remains under researched. The present study investigated how inferencing in the context of three varying concordance lengths (i.e., two complete sentences, one complete sentence, and a truncated sentence) affect EFL learners' accurate inferencing and vocabulary gain. To this end, 66 upper intermediate learners were assigned randomly into three groups and were asked to infer the meaning of 63 unknown words over seven sessions (nine words each session). For each unknown word, three examples in three different lengths were selected. Results indicated that two complete sentence co-text led to more accurate inferencing and vocabulary gain. The pedagogical implications of the findings are discussed.