Speaking is the most complex skill of English Speaking is the most complex skill of English because the speakers need to be focus on some components such as pronunciation, vocabulary and grammar. Moreover, beside those components, pronunciation takes the most crucial not only in English but it also works in other languages. Pronunciation also becomes the biggest problem for students. Everytime English learners in Indonesia speak English, they usually speak English slowly or by pronouncing word by word. Connected speech is the form of spoken English happening naturally or unconsiously to the native speaker in any levels and situations in which can make the English speaker speaks fast and fluently. The purpose of this study is to measure the influence of connected speech toward speaking ability of students at SMP International Islamic Class Program Pamekasan. A pre-experimental design with one group pre-test and post-test was employed in this investigation. In the instrument, the researcher administrated the pre-test at the first time, then the treatment, and finally the post-test. To know the validity of the instruments used, the researcher used content and construct validity in which the test was representative of the material that was given to the students, and for reliability, the researcher used Cronbach’s Alpha by using the SPSS program. This study was done in the ninth grade of SMP International Islamic Class Program (IICP) Pamekasan and consisted of 28 participants. The test score of Ninth grade students was used to collect data for this study. The test was in the form of some questions. The speaking test was conducted to determine the efficacy of connected speech. The test was administrated twice, before and after treatment. The treatment was given before post-test for four times The results demonstrate that after treatment, students' speaking achievement was higher than before treatment. It was proved by the mean score post test was 6.56. while the mean score of pre-test 5.59. To find out whether the difference was significant, the t-test for dependent was applied. The interpretation of the t-test result with df 25, the T-count is 7.086. The 2-tailed significant value is 0.000, which was less than 0.05. The null hypothesis stated that there was no significant influence of students speaking ability after being taught connected speech was rejected. Whereas, the alternative hypothesis was accepted because there was a significant influence of students’ speaking ability after being taught connected speech. Therefore, it is recommended that connecte speech can be applied in teaching speaking skill at Islamic junior high school. Nevertheless, in teaching connected speech, the teacher should be selective in selecting teaching method which suitable to students’ ability in receiving the lesson.