The measurement of the type and number of acid sites on sulfated zirconia catalysts using the P-31 NMR spectrum of adsorbed P(CH3)(3) has been vexed by spectral assignment controversies. Using a combination of NMR experiments and theoretical methods, including chemical shift calculations at the GIAO-MP2 level. we show that a previously observed P-31 resonance at +27 ppm is due to P(CH3)(4)(+), formed in a reaction that consumes a Bronsted site. The coproduct of this reaction. PH(CH3)(2), is protonated on the surface to yield a P-31 resonance in the region expected for P(CH3)(3) On a Lewis site. Further complications result from a signal due to OP(CH3)(3), formed by oxidizing sites on the surface, complexed to unidentified acid sites. As an alternative, we show that carefully designed N-15 experiments using the less reactive and less basic probe pyridine-N-15 provide more easily interpreted measurements of Bronsted and Lewis sites on sulfated zirconias of diverse composition, preparation, and treatment. Quantitative studies revealed that the number of Bronsted sites capable of protonating pyridine corresponded to only similar to7% of the sulfur atoms on the catalyst we studied in the greatest depth. Additional Bronsted sites were created on this catalyst with addition of water, a reaction not observed for sulfur-free zirconia.