The Athens, Greece, earthquake (Mw 5.9) of 7 September 1999 occurred in a region of low seismicity at a distance of ca. 620 km from the epicenter of and only 21 days after the large (Mw 7.5) İzmit, Marmara Sea, earthquake of 17 August 1999. The possibility that the İzmit event triggered the Athens one is examined on the basis of three independent approaches. First, a stochastic test was used, showing that the probability for a strong shock to occur randomly in any 21-day time interval (1) in the region of Athens (radius ∼30 km) and (2) in the broad region of Athens (radius ∼70 km) is of the order of only 5.75 × 10-5 and 2.3 × 10-3, respectively. Second, a model that considers an earthquake triggered by another earthquake as an instability to transient slip was used to calculate the distant shear stress increase at the Athens earthquake fault as a result of the superimposed unrelaxed (coseismic) and relaxed stress change caused by the İzmit earthquake because of a deep aseismic viscous slip. The total stress accumulation within about 1 month of the İzmit earthquake was found to be of the order of 4.2 × 10-3 bars, which is rather significant in that it nearly doubles the yearly mean rate of tectonic stress accumulation in the region. Finally, a time-to-failure analysis showed that the process of small (4.0 ≥ ML ≥ 2.0) earthquake generation, at a distance of about one source dimension (∼30 km) from the Athens earthquake epicenter, started accelerating very slowly from the beginning of 1994. However, only immediately after the İzmit earthquake, the process culminated with short-term foreshock activity. The independent results shown previously favor the conclusion that the Athens earthquake was likely advanced in time possibly because of the stress changes generated by the İzmit large shock.