The purpose of the present study is to determine whether there is long-run relationship between crime rates and unemployment rate in Malaysia for the period 1973 to 2003. The autoregressive distributed lag bounds testing procedure was employed as the main tool to infer cointegration or the long-run relationship between unemployment and the crime rates. The results indicate that the unemployment rate, and crime rates: total crime rate, violent crime (murder, robbery, and assault), and property crime (daylight burglary, night burglary, and motorcycle theft) are cointegrated. The estimated long-run coeffi cients suggest that unemployment rate has negative effect on violent crime, murder, robbery, assault, and motorcycle theft. The paper shows that jobless population in Malaysia as a result of recession tend to remain in or near homes and neighborhoods and this likely will reduce the occurrence of crime.