This study of second language idiom acquisition investigates the interaction of two factors, the L1 effect and the effect of supportive context in the L2 input, when both are available to learners. An experiment consisting of a pretest, a computer-assisted instructional treatment session and a posttest employed three idiom types that differed in terms of shared similarities in first and second languages (Same L1-L2: identical linguistic form and figurative meaning; Different L1-L2: identical linguistic form but different figurative meaning and L2 Only: occurs only in L2), and three task types (a production, an interpretation and a meaning task). On the pretest, participants' scores were the highest for Same L1-L2 type idioms in all tasks. On the posttest, participants' scores were the highest for L2 Only type idioms in all tasks. The analysis of the posttest scores found significant main effects of idiom types and of tasks, indicating a strong treatment effect. This study's findings suggest that the influence of crosslinguistic similarity is lessened when supportive context is available in the instructional input, and that learners can learn L2 idioms, regardless of degree of L1-L2 similarity, when instructional input includes enough context.