We describe here the different tools used for top physics
analysis in the CDF Collaboration. In particular, we discuss how the
jet energy scale, lepton identification, b tagging algorithms and the
neural networks help to improve the signal to background ratio of the top
sample in some cases and to reduce the dominant uncertainties in other.
Results using each one of these tools are also presented.