The performance of surface active additives, such as friction modifiers, depends on their interactions with surfaces. Their effectiveness thus hinges upon the surface conditions. In this work, the effect of cleaning methods of test substrates on the friction reduction capabilities of different organic friction modifier (OFM) additives was investigated. 52100 steel discs and balls were the test specimens. They were cleaned in five different ways. The cleaned surfaces were characterised by using ellipsometry and atomic force microscopy. The tribological performance of stearic acid (STA), octadecylamine (ODA) and octadecanol (ODO) on these surfaces were then tested. As-received steel surfaces were covered with contaminants which may impede the formation of OFM surface layer. Cleaning these surfaces with solvents cannot completely removed these contaminants, with residue layers remain. Cleaning with oxygen or argon plasma results in cleaner surfaces as compared to those cleaned by solvents only. The impact of the choice of cleaning methods on friction depends on the strength of the interaction between the OFM and the steel surface, which determines the ability of an OFM to displace surface contaminations. Cleaner surfaces result in lower initial friction for STA and ODA. Steady state friction is also affected, but to a smaller extent. It may be because most containments remained in the wear track are mechanically removed during rubbing.