Network science approaches have become increasingly popular tools amongst cognitive scientists to model linguistic representations and analyze the effects of individual and group differences on their structural properties. Bilingual semantic networks are concerned with the associations between lexical items in speakers’ first (L1) and second (L2) languages, and existing literature has examined how language dominance and second language immersion impact network organization. Study 1 investigated how differences in L2 proficiency affect the group semantic network structure of Spanish-English bilinguals and English-Spanish bilinguals who are still acquiring their second language. Bilinguals’ L1 networks were more efficient and interconnected than their L2 networks, and increased L2 proficiency resulted in more efficient network organization overall. The goal of Study 3 was to examine the phonemic verbal fluency responses of patients with multiple sclerosis (MS) using a phonological network approach. MS patients consistently exhibit decreased semantic and phonemic verbal fluency performance and have semantic networks that are less efficient and well-organized than aged-matched healthy adults. Despite producing fewer items on the F-A-S phonemic fluency task, MS patients did not produce responses that varied significantly from controls on local phonological network properties, but did produce more low-frequency words that start with “A”. Overall, these studies further our understanding of how linguistic representations are shaped by differences in language proficiency and neurological disease. Their findings also raise methodological considerations that should be addressed when examining differences in language experience via semantic network analyses.