Pronunciation is a factor of two processes: the production and perception of human individual sounds (segments), referred to as phonetics and the combination of these segments in a speech, and referred to as phonology. Ekegusii and Kimeru are Bantu languages spoken in western and central parts of Kenya respectively. University students from the two language groups studying English and Literature in their year one to year four in the university setting formed the population for the study because the intonation and phonetic inventory for both languages are similar to the extent that a non-native speaker of the two languages may not draw a distinction between them. Some prosodic features of these languages such as vowel insertion to break consonant clusters are different from English and when speakers of these languages insert vowels in some English words with consonant clusters, this results in error, sometimes impeding their intelligibility. A study was needed to examine pronunciation errors among the Ekegusii L1 and Kimeru L1 university students so as to document the gravity of the problem. The study was a qualitative description of students’ pronunciation errors in English language committed while the students were participating in university activities. The objectives of the study were to examine the most common mispronounced English phonemes produced by the students and to explore the possible sources of the errors. The study adopted a descriptive study design guided by Corder’s (1974) error analysis model. A purposive sample of 50 students selected on the basis of first encountered first recorded was used for data generation. The data was in the form of notes from listening to the students’ natural talk and audio recordings of their conversations. The study involved describing, analyzing, and interpreting common pronunciation errors. Based on the results of data analysis it was revealed that students made a multiple of pronunciation errors attributable to mainly interlingual and intralingual sources. From the findings, the researchers recommend that the best way to learn the pronunciation of a second language is by listening to good role model speakers of English language and by practising it regularly.