Many students struggle with English fluency due to the use of traditional teaching methods that lack engagement and interactivity. This study aims to assess the impact of the Snowball Throwing model on pupils' speaking abilities at SMPN 34 Surabaya. A quantitative research approach was adopted, employing an experimental design that included both an experimental and a control group. Data collection followed a structured procedure, comprising a pre-test, the application of the treatment, and a post-test using procedure text material. This study used the cluster random sampling approach, which yielded a total sample of 58 pupils. Class VII-I was selected as the experimental group in this case, while Class VII-H was used as the control group. A speaking ability test served as the study tool, and IBM SPSS Statistics version 19 was used to analyze the data. The findings revealed a two-rooted p-value of 0.000, below the significance level of 0.05. Consequently, the null hypothesis (Ho) was rejected, and the alternative hypothesis (Ha) was validated, indicating that the Snowball Throwing model had a noteworthy effect on students' speaking abilities. It may be concluded that using the Snowball Throwing model influences students' speaking abilities because the p-value stayed below the significance threshold.