New palaeomagnetic results are presented for the Pleistocene (MIS2-4) portion of deep-sea core MD98-2181 (MD81; Devao Gulf, Philippine Islands). MD81 is the highest resolution (similar to 50 cm ky-1) palaeomagnetic secular variation (PSV) record for similar to 12-70 ka ever recovered from equatorial latitudes (+/- 15 degrees.). Magnetic studies indicate that MD81 has a stable natural remanence with directional uncertainties (MAD angles) typically less than 3 degrees. We have also recovered a relative palaeointensity estimate from these sediments based on normalization to isothermal remanence. We have correlated our relative palaeointensity record with highresolution relative palaeointensity records from the North Atlantic Ocean. The MD81 ages are always within +/- n 500 yr of the North Atlantic records over the entire core. We also correlate our PSV record with another published PSV record from Indonesia (MD34). We are able to correlate 25 inclination features, 25 declination features and 24 relative palaeointensity features between MD81 and MD34. We identify three intervals of 'anomalous' directions in the cores (based on > 2 sigma deviation from mean directions). One of these intervals contains true excursional directions and is dated to similar to 40.5 ka. We associate this interval with the Laschamp Excursion. We also note two other intervals that have anomalous directions, but no true excursional directions. These intervals occur around similar to 34.5 and similar to 61.5 ka and we associate them with the Mono Lake Excursion (similar to 33.5-34.5 ka) in western USA and the NorwegianGreenland Sea Excursion (similar to 61 +/- 2 ka) in the North Atlantic Ocean. We view our 'anomalous' PSV in the three intervals to be truly anomalous even though most directions are not truly excursional. We think that it is time to reconsider the definition of what is ` anomalous' PSV or excursions. To do that we need good-quality PSV records from several regions that have reproducible records of normal PSV, excursional waveforms and relative palaeointensity. We cannot assess the difference between normal PSV and excursions without such complete PSV records. This study is one attempt to develop such a regional perspective.