The objective of the study is to get the empirical evidence of the answer to the problem formulation; it is whether the English summative test for the eighth grade students of SMP Muhammadiyah 4 in the second semester of the 2008/2009 academic year reliable or not. The writer used parallel test form technique in estimating the reliability. The first test: was administered on June 5th, 2009 and the parallel test was administered on July 13th, 2009. The length of the time slot between the first and the second test administration due to holiday of the school and can be assured that there is not treatment. In collecting the data, the writer chose the eighth grade students.The population of the research is the students of SMP Muhammadiyah 04 Tangerang, the number of whom is 314 students and distributed into 7 classes. Each class consists 44-45 students. The writer used 8A class consisting 45 students as samples and using purposive cluster sampling technique. However, when the test was administered, only 30 students were taking part. The writer used The research finding is that rt = 0, 361 and rh = 0. 98. Because rh > rt, it means that there is a significant correlation between English Summative test made by the teacher scores and the writer made test scores and from the statement above can be said the English Summative Test for Eighth Grade Students of SMP Muhammadiyah 4 Tangerang In The Second Semester of The 2008/2009 Academic Year is reliable.