We use high-precision analyses of fresh basalt glasses from the submarine, Cretaceous Ontong Java Plateau (OJP) recovered by the Ocean Drilling Programs (ODP) to test for long submarine lava flows. Major elements and Cl show that glasses from ODP Holes 1185B and 1186A, which are 200 km apart, are compositionally identical and must have erupted simultaneously from the same well-mixed magma chamber. Identical CO2 contents show they erupted at the same water depth, but 1185B is directly downslope from 1186A and has a corrected basement depth that is >700 m deeper. Remarkably, <2 °C cooling took place over 200 km, requiring a cooling rate <0.01 °C/km.Further evidence of long flows is the lack of vesicles in OJP's glasses, in contrast to mid-ocean ridge basalt glasses that almost always contain vesicles and are often oversaturated in dissolved CO2. We propose that OJP's lavas degassed to saturation levels of dissolved CO2 +H2O and lost all bubbles because they remained liquid for much longer periods during long distance flow. Dissolved CO2+H2O are used to estimate paleoeruption depths which are compared to current corrected basement depths. Using a fixed value of plateau subsidence (500 m) we reconstruct flow distances of 0 to 900 km for eight lava groups from six drill sites. Eruption depths in Hole 807 C are 3040 m for Kwaimbaita-type lavas but 1110 m for Singgalo-type lavas that directly overlie them suggesting that the Singgalo lavas erupted 900 km away (near Site 1183).Paleoeruption depths for Cretaceous glasses from the Jurassic East Mariana and Nauru Basins adjacent to OJP are shallower than most glasses recovered from OJP even though their reconstructed basement depths are 1500–3800 m deeper. It is very likely that the basins’ lavas were erupted on OJP and flowed up to 1600 km: longer than any lava flow known on Earth. Flow distances are shorter (≤1000 km) if a plateau subsidence of 1500 m is used in the calculations. Extremely high Cl in Nauru Basin glasses also supports eruption from OJP. Long distance flow with minimal cooling was facilitated on OJP by favorable slopes, lack of barriers, insulation of rapidly forming glass, and possibly by flow within lava tubes. Lava effusion rates were probably >100 m/sec and flow rapid: >1 m/sec.