This research aims at finding out the effectiveness of using numbered heads together in teaching reading comprehension at the first grade students of SMKN 1 Hu’u in academic year 2015/2016. This research is an Experimental study that quasi experimental design with non-equivalent control group design. The populations of this research were 120 students of 6 classes which termed as X.A (experimental class) and X.B (control class) were selected by cluster random sample technique from the first grade students of SMKN 1 Hu’u in academic year 2015/2016. It showed that the mean scores of the experimental class was 31.5 and the mean scores of control class was 7.2. Moreover, it found that the significance value of t-test > than t-table. It was found that t-table 2.042 at the confidence level 0.05 (95%) and 2.750 at the confidence level 0.01 (99%). After the writer compared that score of t-test is 4.40 > than t-table 2.042 at the confidence level of 0.05 (95%) and 4.40 > than t-table is 2.750 at the confidence level 0.01 (99%). And the degree of freedom (df) is n-2=38 from 120 students as the sample of this research. After analysing the data obtained, the writer concluded that the use of numbered heads together technique is effective in teaching reading comprehension. It indicated that the alternative hypothesis (H3) which is read numbered heads together technique is effective in teaching reading comprehension is accepted.