Abstract The Ngakoringora Ridge is a large (300 m long), linear mound of limestone, dolomitized limestone, and chert that rises abruptly from the desert floor on the southwestern edge of the Lothidok Hills, west of Lake Turkana, Kenya. The origin of the ridge has been controversial. It has previously been considered either a hot-spring deposit or an uplifted holier of pre-Cenozoic marine carbonates. Interpretation has been hampered by the extensive diagenetic alteration of the rocks and the lack of identifiable fossils. A preliminary examination of the ridge morphology and the facies and fabrics of the rocks confirms a hydrothermal origin. The carbonates contain radial calcite fans, micrite-microsparite laminae, stromatolites, coated grains, peloids, intraclasts, Mn-shrubs, and other fabrics that characterize hot-spring travertines. Many of the carbonates have been partly silicified and dolomitized. The ridge is interpreted to be a fissure-ridge travertine, precipitated from thermal waters that discharged along a medial fissure. Fluids flowed laterally from five main mounds that were probably active at different times. Crystal fabrics along the ridge crest are compatible with abiotic precipitation from alkaline spring waters undergoing rapid degassing of CO 2 . Microbially influenced fabrics, including stromatolites, become more common distally. Silicified plants, filamentous microbial mats, and thin chert beds are locally present in distal slope settings. Plant silicification and chert formation may have taken place in shallow terrace pools from spring fluids undergoing cooling and evaporation. The Ngakoringora Ridge formed after the faulting and tilting that formed the Lothidok Hills, but its age is difficult to constrain. Silicification and dolomitization of the carbonates resulted from contact with hydrothermal fluids, and possibly from circulating ground water or lake water after deposition. Hydrothermal activity in rifts migrates with the evolving structural configuration. In the Kenya Rift, this is evident as a migration of hydrothermal activity toward the rift axis. Fossil spring deposits can provide much useful paleoenvironmental information even though they are of small lateral extent.