Background Mental health consequences of disasters are frequently studied. However, few studies have investigated symptom improvement in victims after natural disasters. This study aimed to identify predictors of 6 months post-disaster stress symptoms and to study 6 months and 24 months course of symptoms among Norwegian tourists who experienced the 2004 tsunami. Methods Norwegian tourists (≥18 years) who experienced the 2004 tsunami (n = 2468) were invited to return a postal questionnaire at two points of time. The first data set was collected at 6 months (T1, n = 899) and the second data set at 24 months post-disaster (T2, n = 1180). The population studied consisted of those who responded at both assessments (n = 674). Impact of Event Scale Revised (IES-R) was used to measure posttraumatic stress symptoms. IES–R score ≥33 (caseness) was used to identify various symptom trajectories from T1 to T2. Multiple linear regression was used to determine predictors of posttraumatic stress at T1 and to identify variables associated with symptom improvement from T1 to T2. Results The majority was identified as non-case at both assessments (57.7%), while 20.8% of the respondents were identified as case at both assessments. Symptoms at T1 were positively related to female gender, older age, unemployment, being chased or caught by the waves, witnessing death or suffering, loss of loved ones, experiencing intense fear during the disaster, low conscientiousness, neuroticism and low levels of social support. The IES-R sum score declined from 24.6 (SD = 18.5) at T1 to 22.9 (SD = 18.3) at T2, p < 0.001. Emotional stability and high IES-R scores at T1 were positively related to symptom improvement, while received social support was not. Being referred to a mental health specialist was negatively related to symptom improvement. Conclusions A significant minority (20-30%) among Norwegian tourists developed enduring posttraumatic stress symptoms in the aftermath of the 2004 tsunami. Tsunami exposure, peritraumatic fear, neuroticism and low levels of social support were the strongest predictors of posttraumatic stress at 6 months post-disaster. Decrease in posttraumatic stress was related to emotional stability and higher symptom levels at T1. Being referred to a mental health specialist did not facilitate symptom improvement.