Unfornately, this is hard to do. XRD can only give you an average of the diameter and the accuracy is poor especially when the nanorods are thick. I have a paper that using SEM images and statistical methods to estimate the distribution, and may help you.

it's titled "A Statistics-Guided Approach to Precise Characterization of Nanowire Morphology".  on ACS Nano, 2010, 4, 855.

Can you give more information about the deposition method and the morphology of your nanorods? Is it deposited by CVD or in solution? Are the nanorods dense enough that they form a continuous film, or is there open space between each nanorod?

Thank you so much Dr Fei Wang, Actually, I deposited ZnO dense nanorods using electrochemical technique, the film was continuous without open space . 

you should scan the surface and choose a homogeneous area of your samples, 

you can see more by AFM analysis and determine a lot of parameters  like diameters of grains, clusters, spectral density, ...using software of the AFM.