Eckermann’s (2007) descriptive grammar of Bukawa sheds new light on an important language long overshadowed by its better-documented sibling, Jabêm (also spelled Yabem). Although the syntax and lexicon of the two languages remain quite similar, the phonology of Bukawa now looks to be far more innovative. Newly voiced obstruents in syllable onsets have destroyed the correlation between obstruent voicing and tone that makes the latter much more predictable in Jabêm. Ross (1993) has proposed that the voicing of onsets in Bukawa depended on their metrical position in iambic feet, such that oral stops on weak syllables were voiced, while those on strong syllables were devoiced. Using new data from Eckermann, this review article proposes that another set of voiced obstruents in Bukawa has arisen via postplosion of the nasal onsets of syllables with nonnasal codas. In syllables with nasal codas, by contrast, nasal onsets are preserved and vowels are nasalized, even when final nasal stops are elided. This suggests that oral vowels are already in some degree of phonemic opposition to nasal vowels in Bukawa, just as they are in Kela, the third sibling in the tiny North Huon Gulf subgroup of Oceanic.