Based on worldwide popular novels about Harry Potter written by J. K. Rowling, this study provides an original insight into sentiment distribution in the series based on text analysis. It shows what sentiment appears in the book series and how it changes throughout the chapters. Our analysis confirmed that the Harry Potter book series are more negative than positive. It means that 111 chapters out of 200 have a negative sentiment. Another contribution is finding out what the story is about using text analysis. Our findings confirm that an automated approach to book analysis mostly is in accordance with the reader’s impression. Consequently, natural language processing techniques can help infer what each chapter of the book is about. Keywords: book analysis; Harry Potter; sentiment analysis; text mining; natural language processing