Restricted production of lake trout fry in eastern Lake Ontario limits recruitment and the establishment of a self-sustaining population. To examine the survival and development of eggs and fry, incubation studies were conducted from 1989 to 1993 in situ on Yorkshire Bar and in the laboratory with untreated surface water from Lake on the Mountain. Eggs were collected at spawning time (16 October-9 November, median 24 October) from Lakes Ontario and Manitou, fertilized, and held individually for an average of 162 days in 290 50-cell Plexiglas incubation chambers. On Yorkshire Bar, significantly more eggs died (P < 0.001, 1.65x mean absolute difference (<Delta(x)over bar>) = 14%), and regardless of egg source, there were fewer hatched fry than in the laboratory (P < 0.001, 0.63x, <Delta(x)over bar> = 29%); 75% of the fry that hatched died early in the incubation period, and 42% of these completely decomposed and disappeared by the end of the incubation period, leaving empty incubation cells. Significantly fewer live fry were produced on Yorkshire Bar than in the laboratory (11.2% versus 69.2%, <Delta(x)over bar> = 58%). Abundance of live fry was inversely related to cumulative thermal units (CTU). Increased exposure on the bar from 370 to 690 CTU decreased survival from 25% to 0. In eastern Lake Ontario, spawning has been observed at a mean water temperature of 11.5 degrees C (29 October) but ranged between 12.7 degrees C (19 October) and 8.8 degrees C (14 November); fry survival on 1 May from these dates and temperatures would be 10%, 0%, and 21%, respectively. If the mean temperature at spawning were 2 degrees C lower (9.5 degrees C, 9 days later), production of fry on 1 May would be almost double (19% vs. 10%). Spawning at lower temperatures would increase fry production on shoals like Yorkshire Bar. Temperature inversely affects fry survival, especially if the spawning substrate is degraded by organic sedimentation, which causes increased biological oxygen demand and reduces oxygen concentrations.