A cross-linguistic comparison of syntax acquisition patterns examined preferential subject omission in the naturally-occurring speech of three Japanese toddlers from the Tokyo area and compared the findings to data on English acquisition. Results indicate that acquisition patterns of Japanese do not mirror those of English with regard to this phenomenon, suggesting that the English-speaking child's hypothesis concerning pro-drop parameter-setting is not universal. An alternative explanation concerning configurational, not-pro-drop, parameter-setting and branching direction is proposed. (MSE)