BACKGROUND: This study examined the accuracy of the 17-item Dutch version of the Davidson Trauma Scale (DTS) and the four-item SPAN (Startle, Physiological Arousal, Anger and Numbness) to detect survivors at risk for posttraumatic stress disorder (PTSD) within the first 2 weeks after the trauma. METHODS: 203 civilian survivors of recent trauma with relatively mild symptoms completed the DTS a mean of 8.7 days after experiencing trauma. SPAN scores were computed from the DTS. At a mean of 64.6 days posttrauma, 160 respondents were assessed for diagnosis of PTSD with the Structured Interview for PTSD. RESULTS: Receiver operating characteristic curves showed that the DTS showed good overall screening accuracy (84%). At a cut-off value of 64, the DTS demonstrated a sensitivity of 0.86, a specificity of 0.70, a positive predictive value (PPV) of 0.12, and a negative predictive value (NPV) of 0.98. Overall accuracy of the SPAN was good (89%). At a cut-off of 10 the SPAN showed a sensitivity of 0.86, a specificity of 0.86, a PPV of 0.22, and a NPV of 0.98. The low PPVs were possibly due to the low of prevalence of PTSD in our sample (4.4%). CONCLUSIONS: This study shows that both the DTS and the SPAN are comparably accurate in screening early trauma survivors at risk for developing PTSD. The very brief four-item SPAN may be preferred over the longer 17-item DTS especially in settings in which time and resources are limited. Future studies should aim to cross-validate these results in random samples.