A number of assumptions or requirements must be substantively met in order for the linear regression model parameters to be best linear unbiased estimators (BLUE). That is, parameters must be unbiased, asymptotically efficient, and consistent. This chapter discusses in detail the consequences of violating one or more of the six main regression model assumptions (the disturbance terms have zero mean; the disturbance terms are normally distributed; the regressors and disturbance terms are not correlated; the disturbance terms are homoscedastic; the disturbance terms are not serially correlated (non-autocorrelated); and a linear-in-parameters relationship exists between the dependent variables and the independent variable). Possible corrections are discussed, and the effect that potential violations may have on the interpretation of findings is illustrated through various examples.