Among the signs that humans produce and combine in complex manners to refer to their internal and external worlds, some seem to assume, in a given language and under certain circumstances, an iconic value that others lack, i.e. they seem to associate to their referent in a less arbitrary manner. Among such signs, attention has been drawn to a particular kind of phonemic combinations that generally appear at the beginning or at the end of a lexeme, and which assume iconic connotations not only by the evocative allure of the sounds that make them up, but also, and perhaps primarily, by virtue of being shared among a family of cognate words (e.g. /sl-/ in slide, slip, slither, slime, slobber, slush etc.). We refer to such submorphemic units by the term ‘phonaesthemes’ (Firth 1930). While the objective, universal iconic value of such combinations is, to say the least, moot, it is undeniable that, in the minds of speakers, they are endowed of a certain psychological reality (see Hutchins 1998, Bergen 2004, Pozza and Civale 2022). Germanic languages, and English in particular, display an interesting array of functional and productive phonaesthemes, which have been noticed ever since the 17th century (see Wallis 1653). Besides being an effective means to form, recognize, store and produce vocabulary for the native speaker, it has been suggested that iconicity may be a useful learning tool for young L1 learners (Yoshida 2012, Matsumi and Sotaro 2014) as well as older foreign language learners (Parault and Schwanenflugel 2006, Roig-Marín 2017, Li 2020). In order to assess this latter possibility, however, it is first necessary to test the - spontaneous or induced - psychological reality of such sound-meaning associations in the learners’ minds. This paper reports on a small-scale pilot study conducted on 24 upper-intermediate learners of English from the University of Genoa (Italy), who were assigned vocabulary recognition and memorization tasks based on a list of 78 lexemes belonging to 8 phonaesthetic families, plus a control group of 16 items, in order to investigate whether the presence of phonaesthemes would ease access to the meaning of the target words. The results of the experiment seem to point to a possible exploitation of phonaesthetic families in vocabulary teaching by making the learner notice the correlation between sounds and meanings. However, while an improvement in recognition of the general iconic connotations of the target words can be clearly detected, it remains moot whether the actual similarity between the lexemes within a group could not instead be a hindrance to the discernment of the more specific meanings of the single words.