Purpose: Early apraxia of speech (AOS) research has characterized errors as being variable, resulting in a number of different error types being produced on repeated productions of the same stimuli. Conversely, recent research has uncovered greater consistency in errors, but there are limited data examining sound errors over time (more than one occasion). Furthermore, the influence of conditions of stimulus presentation (blocked vs. random) on sound errors remains uncertain. The purpose of this investigation was to examine the effects of repeated sampling and conditions of stimulus presentation on speech sound errors for 11 speakers with AOS/aphasia. Method: Trisyllabic words consisting of 7 target phonemes in the initial position served as stimuli. On 3 occasions, stimuli were elicited under 2 conditions: blocked (by phoneme) and randomized presentation. Speech productions were analyzed via narrow phonetic transcription. Results: Findings revealed a similar overall mean percentage of errors in both conditions and across sampling occasions. Distortions were the dominant error type. Conclusion: There was no obvious pattern of responding across sampling occasions or conditions of stimulus presentation. The dominant error type differed among target phonemes, but there appeared to be some degree of consistency in the error types produced for the majority of target phonemes. (Contains 4 tables and 3 figures.)