In English, the distinction between voiced and voiceless stops is signaled by a combination of acoustic cues in speakers’ productions. While VOT is the most widely studied, there are many other spectral and temporal cues to the voicing contrast in word initial (“pat/bat”), medial (“rapid/rabid”) and final (“bat/bad”) positions. Recent phonetic studies found that speakers vary in their use of these multiple cues (e.g., Clayards, 2018) and that this individual variability is highly structured: Although the nature of phonetic variation may differ, speakers show consistent covariation of voicing cues within and across phonological categories. Recently, we investigated variability in the use of temporal cues within one voicing category, in word-initial /b/ in connected speech (Jacewicz et al., 2021). We found that closure duration and closure voicing duration in /b/ in adult females varied predictably with dialect and prosodic prominence. However, an unexpected finding was that covariation of these cues also signals a voicing distinction in coda stops (/t/-/d/), which indicates that information about coda voicing can be found not only in the duration of the preceding vowel (longer before voiced than voiceless stops) and characteristics of the coda stops, but also at the onset of a monosyllabic word. This indicates a syllable-wide distribution of cues to stop voicing, reflected in systematic temporal relationships among onset and coda stops. The current study seeks to replicate these results in adult males (n=45, N = 2610 productions) and further examine long-distance covariation of temporal cues to consonant voicing in complex codas (/ts/-/dz/)