The stability of seismicity rate in central and southern Italy was examined in two data sets: from 1975.0 to 1995.0 with M ≧ 3.4 and from 1987.5 to 1996.0 with M ≧ 2.5. These are the approximate minimum magnitudes of complete reporting for the respective periods. The first set was used to evaluate the possibility that the 23 November 1980 Irpinia (M 6.9) earthquake was preceded by precursory seismic quiescence; the second was used to evaluate the conditions under which a current seismic quiescence could be detected in central or southern Italy. During the years before the Irpinia earthquake, the seismicity rate in the northern half of the rupture area and north of it was low. Whether this was a case of precursory quiescence or not is subject to interpretation because the background rate cannot be established for the years before 1975. If we accept the relatively constant seismicity rate in the Irpinia volume during the decade after the mainshock as representative for the background rate, we have a clear case of precursory quiescence lasting at least 1.3 yr up to the mainshock. Alternatively, it can be postulated that the seismicity rate during the decade following this shock was elevated regionally because of the stress redistribution and that the low rates seen before it represent the normal background rate. Even if this reasoning is accepted, the fact remains that a volume including the northern part of the rupture produced no M ≧ 3.4 earthquakes during 1.3 yr before the Irpinia earthquake but produced 10 earthquakes during the 4.7 previous yr. Given these facts, we favor the interpretation that the Irpinia 1980 earthquake was preceded by precursory seismic quiescence. In the entire data set, there are three other cases of quiescence of higher significance without a mainshock following. Since no other mainshock with M > 6 exists in the data set, no missed event exists. We propose that in Italy precursory seismic quiescence may precede mainshocks and that it may be detected in the future by the improved catalog of the modern data set beginning in 1987.5. Major magnitude scale changes give the mistaken appearance that fewer large-magnitude earthquakes occur in Italy now than in years before 1987. We postulate that the rate of earthquakes has not changed and that the magnitude scale should be adjusted.