This　article　reports　a　trajectory　tracking　control　technique　for　snake　robots　with　sideslip　elimination　and　coefficient　approximation.　By　introducing　an　integral　part　and　virtual　input　variables　to　optimize　the　line-of-sight　guidance　law,　a　closed-loop　trajectory　tracking　system　with　the　functions　of　canceling　disturbance　and　sideslip　is　designed.　Besides,　the　method　constructs　the　time-varying　predicted　values　of　virtual　model　variables　and　viscous　friction　coefficients　to　approximate　the　system＇s　unmeasurable　states.　The　approximation　value　can　compensate　for　a　snake　robot＇s　joint　offset　and　torque　input.　Then,　it　is　proved　via　the　Lyapunov　approach　that　the　designed　system　is　stable.　The　remarkable　advantage　of　this　strategy　is　that　the　accuracy　of　a　snake　robot　tracking　the　ideal　trajectory　is　optimized,　which　can　improve　the　error＇s　stability　and　the　body＇s　adaptability　to　the　surroundings.　The　simulation　and　experimental　results　confirm　the　usefulness　of　the　proposed　technique.　©　2005-2012　IEEE.