Few studies have investigated the range and severity of insomnia-related sleep complaints among veterans with posttraumatic stress disorder (PTSD), and the temporal association between insomnia and PTSD severity has yet to be examined. To examine these associations, a large, gender-balanced cohort of veterans (N = 1,649) of the Iraq and Afghanistan conflicts participated in longitudinal assessments of PTSD and insomnia-related symptoms over a period of 2.5 years following enrollment (range: 2-4 years). Data were obtained from multiple sources, including interviews, self-report assessments, and electronic medical record data. Three-fourths (74.0%) of veterans with PTSD diagnoses at Time 1 (T1) reported insomnia-related sleep difficulties on at least half the nights during the past 30 days, and one-third of participants had received a prescription for a sedative-hypnotic drug in the past year. Veterans without PTSD had fewer sleep problems overall, although the prevalence of sleep problems was high among all study participants. In longitudinal, cross-lagged panel models, the frequency of sleep problems at T1 independently predicted increases in PTSD severity at Time 2 (T2), B = 0.27, p < .001, after controlling for gender and relevant comorbidities. Conversely, T1 PTSD severity was associated with increasing sleep complaints at T2 but to a lesser degree, B = 0.04, p < .001. Moderately high rates of sedative-hypnotic use were seen in veterans with PTSD, with more frequent use in women compared to men (40.4% vs. 35.0%). Sleep complaints were highly prevalent overall and highlight the need for increased clinical focus on this area.