An ecological study of spontaneous fermentation in red wine carried out in two wineries located in the same village of the D.O.Ca. Rioja appellation (Spain). These two wineries were very different: one was very old and the other commenced its operations with the 2000 vintage. The study was conducted over seven consecutive years (1997-2003); the first 3 years in the old winery; and the other 4 years in the new one. Another difference between the wineries was the use of commercial strains: no starters were ever used in the old winery and two known strains (A and B) were employed only in a white must vat in the new winery. The number of different Saccharomyces cerevisiae strains detected for each vintage and the frequency of their appearance varied between the old winery and the new winery. There was no one clearly dominant clone in the fermentations studied in the old winery, whereas in the vinifications monitored in the new winery, a clear majority strain was detected. The same S. cerevisiae clone took over the spontaneous fermentation of all the vintages studied in this new winery. Analysis of the mtDNA restriction pattern of this strain revealed that it was identical to one of the commercial strain used (starter A), despite the fact that the fermentation studied was spontaneous. Microbial analysis conducted on winery equipment and winery surfaces prior to the 2003 vintage revealed that starter A had become a resident of the winery with a percentage of 50% among all S. cerevisiae. The presence of non-Saccharomyces yeasts in the early stages of vinification also differed between the old and new wineries.