In an attempt to study variations in perceived quality as an effect of temporally differing Swedish with a foreign accent, recordings of one Spanish and one Estonian speaker were presented to groups of native Swedish listeners. Both inter-speaker differences as well as intraspeaker differences between intact and manipulated versions were studied. Temporal manipulations refer only to the durations of phonologically long segments. Segments were lengthened in the case of the Spanish speaker and shortened in the case of the Estonian speaker. Listeners tended to rate the pronunciation of the Estonian speaker higher than that of the Spanish speaker regardless of intact or manipulated version, whereas listeners who compared intact version to manipulated version tended to rate versions with “long sounds” higher than versions with “short sounds”. A substantial part of the native Swedish listeners rated intact and manipulated versions equally, but if they rated one over the other, the “long sound ” version was always rated as bestpronunciation with respect to three quality-related variables.