Social media (SM) use has been increasingly recognized as a potential contributor to poor sleep. Few studies have examined SM use and sleep using ecological momentary assessment (EMA), compared different types of media use (SM, television, gaming), or examined whether youth at high and low familial risk for depression are differentially affected by SM use. The current study included 76 youth (46% female; Mean age = 11.28 years) who were recruited based on parental history of recurrent depression (N = 35 high risk; N = 41 low risk) in the United States. Youth completed a 9-day EMA protocol, which included current activity at time of prompt and daily sleep onset and offset times. Regression and multilevel models were conducted to examine the effects of media use on sleep. Results indicated that youth who used more SM (mean and number of days) went to sleep later, but did not have shorter sleep duration. Youth with more SM use also had higher levels of variability of both sleep timing and sleep duration across the 9-day period. There were no effects of gaming or TV on sleep, and youth at high risk for depression did not have differences in SM use or its effects on sleep compared to low-risk youth. These findings indicate a unique impact of SM use on sleep timing and variability for youth (regardless of risk status), which may suggest a unique and modifiable pathway through which SM use contributes to poor health. Copyright © 2020 The Foundation for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.