The study aimed to investigate the effect and efficiency of a training porogram on the social skills of students with learning disabilities The sample was selected by the simple random method. It included 30 male and 30 female students with learning disabilities. The experiment was carried out in schools in Salt Educational Directorate. The sample was divided into two groups (control group and experimental group). Students in the experimental group were exposed to a training program in social skills, while students in the control group were not. The improvement in the social skills was measured by the (Social Skills Test) that was developed by Gresham & Elliot,( 1990). The researcher used the translated form of this test that was developed by (Yusuf, 1992) whose validity and reliability indicators served the purpose of this study. The analysis of covariance (ANCOVA) was used to examine the hypotheses of the study. The study revealed the following results. - In the post test, there was a statistically significant differences (≤0.05) in the social skills between the control and experimental groups. This difference was in favor of the experimental group. - In the follow–up test, there was a statistically significant difference (≤0.05) in the social skills between the control and experimental group. This difference was in favor of the experimental group. - In the post test, there was no statistically significant difference (≤0.05) in the social skills attributed to gender or group or to interaction between them.