The present study investigates the influence of first language (L1) lexicalization patterns on the processing of second language (L2) words in sentential contexts by advanced German learners of English. The focus was on cases where a polysemous word in the L1 is realized by independent words in the L2, e.g. German "Blase" realized by English "bubble" and "blister." An anomaly detection task was used in which participants had to indicate whether a target word formed an acceptable completion to a sentence. The critical condition was where the other sense ("blister") of the translation equivalent Blase was appropriate, but the word ("bubble") did not complete the sentence meaningfully, e.g. "His shoes were uncomfortable due to a bubble." This was compared to a control condition in which neither sense of the L1 translation made sense, e.g. "She was very hungry because of a bubble." Factors of word type (noun vs. verb) and degree of relatedness of L1 senses (high vs. moderate) were also manipulated. Relative to native speakers of English, advanced German learners made more errors and displayed longer correct response times in the critical condition compared to the control condition. An effect of meaning relatedness was obtained for nouns but not verbs. The results are discussed in terms of the role of lexical-level translation connections in activating L1 concepts from L2 words, even in highly proficient learners and in all-L2 tasks. (Contains 3 tables, 2 figures, and 1 footnote.)