skill. Task-Based /..anguag,i t.eorning (Tl/II). Task: teachers' guides. • • • ., A1<1st Ethiopian preparato1J' school students find it velJ' challenging to communicate in Engtish in or out of the classroom. The° objective of this swdy was to investtgate the effect of TBLL program on-devctoping the speaking skills ()( grade I I students at Tagel preparatory school in Simada Woreda during a period offive months in when this study was conducted ( FebruaryJune) of the academic year 20!5!16.The present study auempted to answer the following questtons: First, Does students • speaking performance improve as a result of exposure to TBLL'! {j' so, is the change statistically significant? Second. what cm? the students · perceptions of' the TELL in speaking classes? If is a type of pre-post-test non-equivalent quasi-experimental design . The parl!cipan1s in the study are 9 J eleventh grade students. The researcher selected two classes from 9 ctossesfor the study by using convenient sampling and by drawing a lot section "A" (50) and section 'B' (41) students assigned as experimental and control group respectively. A taskbased language teaching program was developed by the researcherfor the expari1111?1110/ group. Instruments were used in tbis study: I. a pre-post-test and delayed post-test ofspeaking skills to measure the effect of the instructional program 011 developing thr students' speaking skill. 711e tests coC,red the dimensions of accuracy 1111d .. fl11e11cy. The table of specification was used to assess the students' performance he/ore and ofter implememing the designed ;>rogro,n. 2 .. An attitudinal questionnaire aimed at invesiigatinK the experimental students' auitudes before and ofter the implomentution ,i/ the program. 'the validity and reliabitity of the research instruments were checked. 3. A focu« grout: discussion wu:, held to measure the attitudes of experimental students towards task-based language learning. Paired sample t-test and Independent sample 1- 11,;;1 were ILY<«l 10 analyze the quantitative data, and the data goshered 1/mwgh qualitmive data analyzed by 111e researcher qualitatively. The findings of the study were the following: Firstly, the TBLT program sign(/h.-cmliy enhanced lite speaking skit! ofthe students of the experimental group ofter tile treatment. Secondly. the TBLT program has u :,ignific,ml e_{li:ct or/ s1ud,1111s' attitude towards task-based language learning after the intervention. Based on the results of the , . current study, if is recommended that EFL teachers should 1.1dop1 the TBf.T principles and , procedures in their· classroom practices. Finally. the researcher recommended that curriculum designers should incorporate nu;r principles and procedures in she students' books and