We investigate the constraints on the Cardassian model using the recent results from the Wilkinson microwave anisotropy probe for the locations of the peaks of the cosmic microwave background (CMB) anisotropy spectrum. We find that the model is consistent with the recent observational data for a certain range of the model parameter n and the cosmological parameters. We find that the Cardassian model is favored compared to the LambdaCDM model for a higher spectral index (n(2) approximate to 1) together with a lower value of the Hubble parameter h (h less than or equal to 0.71). But for smaller values of n(s), both ACDM and Cardassian models are equally favored. Also, irrespective of supernova constraints, CMB data alone predict the current acceleration of the Universe in this model. We have also studied the constraint on sigma(8), the rms density fluctuations at the 8h(-1) Mpc scale.