Speckle noises appear in synthetic aperture radar（SAR）imagery because of coherent superposition of elec- tromagnetic wave scattering from scattering points in a resolution cell.Multi-look processing is usually used to reduce noise and thus exerts an impact on SAR ship detection. Single-look SAR images were processed with multi-look method to get multi-look images.By analyzing the equivalent number of looks

spatial resolution and ship-sea contrast

we got following results:on the one hand

multi-look processing increases equivalent number of looks and improves the image quality

and this is helpful for ship detection.On the other hand

multi-look processing reduces the image spatial resolution and ship-sea contrast

and this is the disadvantage for ship detection. Ship detection were carried out as case studies for C and L band HH

HV

and VV polarization SIR-C SAR images with different looks by using two typical ship detection methods such as windows filter method and K-distribution method. And Figure of Merit（FOM）was used to measure the quality of ship detection.It is concluded that: （1）the quality of ship detection with less multi-look processing is better than that with more multi-look processing or no multi-look processing（remain single-look）; （2）the quality of ship detection is the best when the number of multi-look processing is 3 for windows filter method ; （3）the quality of ship detection is the best when the number of multi-look processing is 2 or 3 for K-distribution method.