Modeling the continental lithosphere's physical properties, especially its depth extent, must be done within a self-consistent petrological-geophysical framework; modeling using only one or two data types may easily lead to inconsistencies and erroneous interpretations. Using the LitMod approach for hypothesis testing and first-order modeling, we show how assumptions made about crustal information and the probable compositions of the lithospheric and sub-lithospheric mantle affect particular observables, particularly especially surface topographic elevation. The critical crustal parameter is density, leading to ca. 600m error in topography for 50kgm-3 imprecision. The next key parameter is crustal thickness, and uncertainties in its definition lead to around ca. 4km uncertainty in LAB for every 1km of variation in Moho depth. Possible errors in the other assumed crustal parameters introduce a few kilometers of uncertainty in the depth to the LAB.We use Ireland as a natural laboratory to demonstrate the approach. From first-order arguments and given reasonable assumptions, a topographic elevation in the range of 50-100. m, which is the average across Ireland, requires that the lithosphere-asthenosphere boundary (LAB) beneath most of Ireland must lie in the range 90-115. km. A somewhat shallower (to 85. km) LAB is permitted, but the crust must be thinned (<. 29. km) to compensate.The observations, especially topography, are inconsistent with suggestions, based on interpretation of S-to-P receiver functions, that the LAB thins from 85. km in southern Ireland to 55. km in central northern Ireland over a distance of <. 150. km. Such a thin lithosphere would result in over 1000. m of uplift, and such rapid thinning by 30. km over less than 150. km would yield significant north-south variations in topographic elevation, Bouguer anomaly, and geoid height, none of which are observed. Even juxtaposing the most extreme probable depleted composition for the lithospheric mantle beneath southern Ireland against the most extreme fertile composition beneath northern Ireland only allows some 20. km of LAB variation; any further variations would produce effects that are well beyond those observed.One model that satisfies almost all the extant data to first order includes a spinel-peridotite upper lithospheric mantle layer to 85. km in southern Ireland and to 55. km in northern Ireland, thinning over a lateral distance of 150. km. Below this in southern Ireland is a garnet peridotite layer extending down to 115. km, and in northern Ireland a refertilized layer down to 95. km. The mid-lithospheric chemical discontinuity (MLD) at the base of the Spinel Peridotite zone may explain the observed discontinuity in S-to-P (Sp) receiver functions.