We used a typing task to measure the written production of compounds, pseudocompounds, and monomorphemic words on a letter-by-letter basis to determine whether written production (as measured by interletter typing speed) was affected by morphemic structure and semantic transparency of the constituents. Semantic transparency was analyzed using a dichotomous classification (opaque vs. transparent) as well as participant ratings. Our results indicate that written production is sensitive to morphemic structure and to the semantic transparency of the first constituent.