In central California, periodic earthquake occurrence suggests a relationship with annual hydrological, atmospheric, thermal, and tidal loadings. In catalogs of declustered earthquakes within 100 km from Parkfield, CA, we study a semiannual periodicity for the monthly number of ≤7.2-kmdeep earthquakes and the monthly median hypocenter depth over 1994–2002 and 2006–2014. Peaktrough months in fitted periodic components differ between time spans and event populations. Deeper earthquakes present no semiannual and a weak annual periodicity. Although modeled pore pressure shows a Spring peak, when added to the elastic Coulomb stress from surface hydrospheric loads, it fails to predict a 6-month periodicity for 2006–2014. In 1994–2002, the pore-pressure amplitude appears to be of same order as the elastic stress and may have had a stronger effect. In 2006–2014, we did not find load model parameters explaining the observed pattern or apparent changes following the 2003 San Simeon and 2004 Parkfield earthquakes. Plain Language Summary Seasonal surface water loading and other periodic sources of stress changes in the Earth's crust have been shown to modulate seismicity. We evaluate the annual variation in seismicity and its semiannual component of 6 instead of 12 months as a function of hypocenter depth near Parkfield, CA. Two 9-year intervals that avoid the M6.6 San Simeon and M6.0 Parkfield earthquakes and their aftershock sequences are examined: January 1, 1994–December 31, 2002 and January 1, 2006–December 31, 2014. Each of the two populations of events is divided into two equalsized subpopulations: hypocenter depth ≤ 7.2 km and 7.2 km < hypocenter depth ≤ 15 km. Earthquakes exhibit different periodicities in monthly numbers over 1994–2002 and 2006–2014, indicating a changed distribution of earthquake occurrence within the year after 2003–2005, which may be caused by a weakened Spring peak of pore pressure or altered permeability conditions due to the recent mainshocks.