We examine factors that influence accident proneness among employees. We ag
ree that the determinants of accident proneness include organizational, emo
tional and personal factors. Using logistic regression we estimated three m
odels, and their predictability for accident proneness among sample of 200
injured workers interviewed upon entering hospital emergency wards in Israe
l. Work injuries were not contingent on age, religion, nor education. The e
ffects of gender were strong but non-significant. Subcontracted and higher-
paid workers are more likely to get repeat injuries. Prior injury experienc
e sensitized employees to stronger perceptions of risk associated with unsa
fe practices. Large family households, ameliorates stress feelings and less
ens the likelihood of accident proneness while poor housing conditions have
the opposite effect. The full model demonstrates considerable prediction o
f injuries when focusing on type of employment, personal income level, bein
g involved in dangerous jobs, emotional distress and a poor housing environ
ment. The model contains most of the significant results of interest and pr
ovides a high level of predictability for work injuries. (C) 2000 Elsevier
Science Ltd. All rights reserved.