The earthquake prediction model of Whiteside and Ben-Zion [1995> was examined to ascertain whether or not their proposed method generates statistically significant results. Their model suggests a remarkable increase in the triggering of small earthquakes by free-earth oscillations before some large events in the vicinity of the mainshock months prior to the large event. We independently devised a program to reproduce their method and found the results to be dependent upon the number of events used in the generation of triggering percentages. Results generated using real and random data were nearly identical. We devised a parameter to remove sampling bias and test the null hypothesis, i.e., that the results using real data are no different from what is expected for random chance. The null hypothesis is upheld and there is no observed increase in this parameter in the location of the Loma Prieta earthquake prior to that event. ¿ 1999 American Geophysical Union