In phonological conditioning, phonemes become assimilated to match the properties of adjacent phonemes. At the syllable coda morpheme boundary in the English language, these morphemic elements express grammatical features of marking number (plural) and genitive ( singular and plural possession ) in regular nouns, and the 3rd person singular and past tense in regular verbs. These morphemes create consonant clusters that are expected to assimilate for easy production (or linguistic economy). Unfortunately, it has been revealed that non-native speakers do not always assimilate the consonants to reflect this conditioning. This study investigates the (non)assimilation of consonant clusters in phonological conditioning from an acoustic perspective. It uncovers the patterns of phonological conditioning that are derived from (non)assimilation of consonant clusters in Ghanaian English and the factors that account for such patterns using the Speech Learning Model (SLM) by Flege (1995) and the Markedness Differential Hypothesis by Ekman (1977). Six (6) hours of data recorded from 50 Ghanaians were subjected to auditory and acoustic analyses. Results indicate that Ghanaian speakers of English sometimes assimilate these phonological endings correctly but mostly deviate from inner circle expectations. Typically, while they assimilate the voiceless endings, the voiced endings most often pose a challenge to them. Keywords: assimilation, consonant clusters, phonological conditioning, morphophonemic orthography, markedness