In this 1992 project it reanalyzed the data collected by Pate and Fridell ( 1993) on citizen complaints against police use of excessive force. The current report includes two empirical studies on the citizen complaints about police brutality in two mutually excluding areas: the police use of excessive physical force and the police use of all other non-physical forces, such as abuse of authority and verbal abuse. It attempts to establish the baseline correlation of citizen complaint rates with various police organizational factors, and to identify the causal effect of police brutality. Using Tobit regression technique, the research tested a series of hypotheses deduced from theories advanced by Wilson (1968) and Lundman (1980) with a number of control variables. It is found that organizational behavior and organizational characteristics are important covariates of the citizen complaints against police use of excessive physical force and police abuse of power. The police administration can influence its officer's behavior by strengthening the in-service training, paying attention to the education achievements of its officers, and actively provide best training for qualified new police in the force.