This corpus-driven study explores the linguistics phenomenon of mental verbs in English song lyrics from 1960s until 2000s. This study aims to identify the frequency distributions of lexical verbs, mental verbs, and to analyze the language uses of mental verbs in the Diachronic Corpus of English Song Lyrics (DCOESL). Quantitative and qualitative methods of analysis were applied. First, quantitative data covering frequency distributions of general verbs was produced via LancsBox. Top three mental verbs in song lyrics were selected for analysis and discussion. The frequency distributions of mental verbs and collocations were produced via LancsBox. Collocational patterns were illustrated through collocational graphs constructed via LancsBox. Frequency distributions of mental verbs were compared to reference corpus Contemporary Corpus of American English (COCA) for the purpose of generalizing the findings from this study as representative of English language. The statistical data were submitted for four statistical tests of significance namely Chi-square, Mutual Information, Loglikelihood, and t-score. Second, qualitative data was composed of corpus annotations. Corpus annotations were conducted via CLAWS for assigning part-of-speech C7 tagset to identify verbs. Semantic categories of mental verbs were identified via UCREL Semantic Analysis System (USAS). Findings uncovered significantly high frequencies of mental verbs "know," "want," and "love" in English song lyrics through 1960s until 1990s. These three mental verbs possess high inclination to occur alongside personal pronouns "I" and "you," depict social actions, high predilection for simple present tense, and simple sentence structure. These attributes illuminate that song lyrics emulate spoken English, predominantly the informal conversation register.