This paper focuses on the development of tones in Baima (ISO-639 code bqh), a little-studied Tibetic language of the Chinese-Qiang-Tibetan-rGyalrong borderland. Previous studies of the historical phonology of Baima (Zhang 1994a,b; Huang & Zhang 1995) provide an extensive account of the development of Baima initials and rhymes, but devote little attention to the development of its tones. A brief outline in Huang & Zhang (1995: 101-102) concludes that tonal development is conditioned by onset voicing and rhyme length, while leaving many developments unexplained. This paper reexamines this conclusion in light of newly collected data and an updated phonological analysis. The analysis recognizes (i) epiglottalization as secondary articulation in Baima consonants (epiglottalized consonants develop from OT voiced obstruents and sonorants with preradicals); and (ii) phonation as one of the phonetic correlates of tone in Baima (Chirkova et al. forthcoming). In this new analysis, Baima is argued to have a tonal system, characterized by syllable-level contrasts, involving the redundant use of pitch, voice quality, vowel length, and syllable onset co-occurrence patterns. Building on this new analysis, Baima tonal history is argued to have begun with the genesis of the low register that was conditioned by, on the one hand, phonation and laryngeal specification of onset consonants and, on the other hand, rhyme length. The resulting low versus non-low register system was later restructured, following a split of the non-low register, conditioned by the coda type: falling tone in syllables with zero or stop codas, and level tone in syllables with nasal and liquid codas or coalesced syllables. The paper provides detailed arguments for these changes with particular attention to their underlying phonetic mechanisms (including coarticulatory effects of aspirated onsets on the voice quality of following vowels, Garellek & Keating 2011; and interaction between phonation and nasality, Garellek et al. 2016). The paper further comments on ongoing change in Baima contrastive phonation categories, leading to pitch and/or vowel quality overtaking phonation as the primary cue of the original contrast (cf. Keating et al. 2011; Kuang & Cui 2018). The final part of the paper links its findings to broader developments in the neighboring languages (Evans 2006, Evans et al. 2016; Sun 1997, 2001; Sun et al. 2017; Suzuki 2015; Sims 2020).