This study aimed to collect empirical studies related to technology enhanced language learning, and to systematically synthesize and review these studies to investigate the effectiveness of technology on language learning. The variables are year of publication, educational level, sample size, treatment duration, outcome measurement and types of devices. A total of 90 studies published between 1990 and 2015 were comprised in this meta-analysis. For the distribution of the studies over years, more than half of the included studies were conducted during 2011 to 2015 and around 26.7% of the studies were conducted during 2006 to 2010, only 4% were from 1996 to 2000 and 2% were from 1991 to 1995. The overall effect size of the 90 TELL studies is 0.847. The Q test shows that there was a significant difference among these studies, therefore, further investigation on moderators was conducted. The findings reveals that among the moderator variables, learners' educational level, sample size, and learning devices, were found to be significant, and could affect the effectiveness of the technology intervention.