In classical language teaching methods, the weight has first been given to the grammar, and the acquisition of the lexical items used in that language has been relegated to the secondary position as by- products to be gained through the learning of grammar rules. However, as Wilkins (1972) points out, it is possible to express very little without grammar, but without words, nothing, even the items relating grammar, can be conveyed. Since this fact has been presented by many researchers with emphasis on the necessity of vocabulary knowledge, the importance given to vocabulary knowledge has increased in the language teaching, and also language teaching methods developed from the words, especially the content words, frequently used in any language have gained momentum since the last quarter of the 20th century. The developed language teaching methods emphasize the importance of the fixed expressions made up of the words which intensively co-occur rather than individual words. In many of the earlier studies, there are explanations concerning the benefit fixed expressions provide in encoding and decoding the new language learners try to acquire. The results of computer-based studies have also shown that formulaic language is used extensively in the sentences and texts produced by native speakers (Erman and Warren, 2000, Schmitt and Carter, 2004: 1). In view of these explanations, this study examined to what extent fixed phrases were used by learners of Turkish and which fixed phrases were preferred most frequently. The data used in the study came from a small-scale corpus generated from two compositions written by 45 Swahili speakers who came to Turkey to attend the School of Medicine in different universities. Before their education, they had to learn Turkish for an academic year in the preparatory course offered by Turkish Learning and Research Center (TÖMER) in Gazi University. If they were successful in the exam at the end of the year, they could start their education. After collecting the data, the commonest fixed phrases occurring in the compositions of the participants were identified and their influence was investigated on the effective and fluent use of the language production. The results obtained from both the quantitative and qualitative analyses showed that Swahili speakers learning Turkish made use of fixed phrases in their written production. However, they were restricted in number and variety. In view of the results, the place of fixed phrases in Turkish teaching was evaluated and the methods and techniques that can be applied to teach them effectively were discussed.