This study builds on research investigating the construct validity of elicited imitation (EI) as a measure of implicit second language (L2) grammatical knowledge. It differs from previous studies in that the EI task focuses on a single grammatical feature and time on task is strictly controlled. Seventy-three EFL learners and 20 native English speakers completed the EI and four other tests hypothesized as measures of implicit or explicit L2 knowledge. Factor analytic results indicated that learners' EI scores loaded on the factor labeled implicit L2 knowledge, confirming previous findings. Results from other tests and methodological issues concerning EI design and use suggest that the construct validation of EI as a measure of implicit L2 grammatical knowledge awaits further investigation.