MALDI-TOF MS-based yeast identification requires a pretreatment step for which four are described in the literature, i.e., direct smear, fast formic acid and two complete formic acid/acetonitrile extractions. In this study we compared the impact of these procedures on the performance of MALDI-TOF MS-based yeast identification of samples from colonies grown on Sabouraud or chromogenic media. A total of 103 yeast isolates recovered from clinical samples were identified in parallel using the four pretreatment procedures. The proportions of both correct identifications (regardless of LogScore values) and of reliable identifications (i.e., correct identifications with a LogScore 2, as recommended by the manufacturer) obtained with the four techniques were compared. Even if the proportion of correct identifications exceeded 85% independent of the pretreatment procedure, results obtained with complete formic acid/acetonitril extractions of colonies grown on Sabouraud media were significantly superior to those with smear and fast formic acid extraction procedures. If one considers only reliable identifications, then both smear and fast formic acid extraction procedures yielded lower (<40%) correct identification rates than the use of the two complete extraction procedures (>77%) of portions of colonies on both Sabouraud and chromogenic media. The data would indicate that the direct smear and fast formic acid procedures cannot be recommended due to the LogScore values which were continually below those recommended by the manufacturer for biological validation. Thus, complete extraction methods are better suited for MALDI-TOF MS-based yeast identification in the clinical laboratory setting although they are more labor-intensive.