Oceanic microseisms are generated by the interaction of opposing ocean waves and subsequent coupling with the seabed, so microseisms should contain information on the ocean conditions that generated them. This leads to the possibility of using seismic records as a proxy for the ocean gravity wavefield. Here we investigate the P-wave component of microseisms, which has previously been linked to areas of high wave interaction intensity in mid-ocean regions. We compare modeled P-wave microseismic sources with those observed at an array in California, and also investigate the relationship between observed sources and significant wave height. We found that the time-varying location of microseism sources in the North Pacific, mapped from beamforming and backprojection of seismic data, was accurate to ≤10° in 90% of cases. The modeled sources were found to dominate at ∼0.2 Hz which was also reflected in the seismic observations. An empirical relationship between observed beampower and modeled source power allowed sources during an independent data period to be estimated with a correlation coefficient of 0.63. Likewise, significant wave height was also estimated with a correlation coefficient of 0.63. Our findings suggest that with improvements in resolution and amplitude retrieval from beamforming, correlations up to 0.78 should be possible between observed P-wave microseisms and significant wave height in remote ocean regions.