Abstract Nurses constitute the largest group in the healthcare workforce and are called on to assist in emergencies such as disasters. Research has shown that professionals with higher levels of knowledge are more likely to respond to actual emergencies. Yet most hospital based nurses do not possess the skills needed for disaster response. The Basic Disaster Life Support (BDLS) course, with its comprehensive content, represents the gold standard for disaster education. Since confidence also plays a role in response, a tool to measure this variable could be useful. There were five purposes of this study: determine whether one teaching method (computer or classroom instructor-led) is superior over another for disaster education; evaluate how knowledge retention varies between instructional models; examine whether a correlation exists between self-efficacy and disaster knowledge; pilot a new instrument, Disaster Self-Efficacy Scale (DSES); complete psychometrics on the Basic Disaster Life Support exam. The study was an experimental pretest/posttest/follow-up with a single between-group factor (type of training with three levels) and three within-group factors measured at three intervals. The sample included 82 hospital-based nurses randomly assigned to a computer-based, instructor-led, or control group. A MANOVA and MANCOVA were conducted to evaluate group differences at three time intervals. Psychometric evaluation was conducted on both the BDLS and the piloted Disaster Self-Efficacy measures. The BDLS test was shown to be in need of revisions and updating. The DSES measure shows promise for determining disaster self-efficacy and may be useful to target training though it needs further validation. Learning results showed that when controlling for pretest differences, experimental groups had higher posttest BDLS and DSES scores than the control group but there was no difference between experimental groups. There was no difference between experimental groups for BDLS scores at follow-up. Conclusions were that training, regardless of how it was delivered, led to a dramatic increase in disaster knowledge and disaster self-efficacy; computer-based education is a feasible alternative to teaching BDLS; retention still poses a challenge for disaster education. Implications for nursing education and practice were identified. Future research should focus on further development and validation of the DSES and BDLS instruments.