This study analyzes the Stop and Frisk policy in New York City to see if there is any evidence of racial profiling in force used by police officers. The study uses a logit model to estimate the probability of having force used on an individual conditional on being stopped. The probability of having force used on blacks and Hispanics is higher. A model specification with an interaction term between race of the suspect and crime rate of the precinct where the individual stopped further suggests that the difference in probability of force between blacks, Hispanics and whites is highest in low crime rate precincts. For precincts with higher crime rates the difference in probability between blacks, Hispanic and whites is lower.