The current study aimed at investigating the effectiveness of a training program based on features of connected speech in an adaptive learning environment in listening skills among EFL majors. To meet such an end, the randomized pretest-posttest control group design was adopted. Participants, totaling 66 third year EFL majors at the Faculty of Education (Cairo), Al-Azhar University, were randomly selected and assigned into an experimental group (N=33) receiving the experimental treatment and a control one (N=33) receiving no instruction. Consequently, the researcher developed a listening skills checklist and a listening skills test which was administered after verifying its validity and reliability. Moreover, an adaptive learning environment was developed according to the level of proficiency, low, mid and high. More specifically, the researcher developed, firstly, the training program which comprised ten features of connected speech for the three levels, secondly, three scenarios, one for each level, to be transformed digitally by programmers. The results of the independent sample t-test revealed the outperformance of the experimental group students (mean=24.39) over those of the control one (mean=13.60). Furthermore, Eta squared (2) results showed a large effect size (.560) on the listening skills of the experimental group students. Such results indicated that the experimental treatment was effective in developing the third year EFL majors' listening skills and that 56% of the effect size could be attributed to the training program. The study recommended the need for instructing the features of connected speech in adaptive learning environments to develop the listening skills among EFL majors and suggested designing such environments according to more variables containing more features of connected speech and targeting different participants.