Purpose: This study was done to develop a prediction model for drug misuse and abuse behavior by elders living in the community. Methods: Based on cross-sectional research, a covariance structure analysis was performed to predict the paths among the factors influencing drug misuse and abuse behavior in elders. The participants in this study were 357 elderly people. Data were analyzed using SAS 9.1 and LISREL 8.53 to test the model. Results: The fitness indices for the modified model were x2=2.371 (p= .796), GFI=0.997, SRMR=.018, AGFI= 0.989, NFI=0.988, NNFI=1.042, CFI=1.000, and IFI=1.013. Drug misuse and abuse behavior were directly influenced by attitudes toward drug use, and influenced, both directly and indirectly, by depression and uncertainty in the model. However, knowledge about drug use and social support while being directly influenced by depression and uncertainty did not influence drug misuse and abuse behaviors in the model. Conclusion: Our findings suggest that to prevent drug misuse and abuse behaviors among elders, health professionals must thoroughly assess depression, uncertainty, and attitudes toward drug use.