There are many earthquake catalogs covering California, but none includes everything one would like to have for constructing and testing hypotheses of earthquake occurrence. Perhaps it is most obvious that focal mechanism estimates are needed in many phases of earthquake forecasting and evaluation. They are needed for estimating the stress increments from past earthquakes, for identifying those earthquakes that have been encouraged by known previous earthquakes, and for testing models that forecast focal mechanisms as well as location and magnitude. Focal mechanisms are available for recent, instrumentally recorded earthquakes, but they are not yet uniformly available for all relevant quakes, and there are often alternate solutions for recent events. Some focal mechanisms are available for large historic (that is, pre-instrumental) earthquakes, but these estimates are generally scattered through the literature rather than systematically cataloged. Instrumentally determined focal mechanisms are clearly best, but estimated ones can still be useful. Measured focal mechanisms show some degree of regularity that can be exploited to estimate unmeasured ones from nearby measured values. Some earthquake forecasts explicitly apply to mainshocks as well as aftershocks, while others, like the source model for the 2002 National Seismic Hazard Map, are meant to apply to mainshocks only. For those reasons it would be convenient to have a catalog in which all known events are included, but which labels aftershocks in a way that they can be removed or down-weighted as needed. Earthquake catalog data, like any other results, are subject to errors. Location errors for historic earthquakes and magnitude errors for all events could well affect the assessment of an earthquake forecast. Statistical tests can and should account for the uncertainties in catalog data. To date, such tests have not been employed for earthquake forecasts, in part because the uncertainties are poorly understood. It is well known that …