The liquefaction potential assessment under the framework of simpliˆed procedure that involves the use of cone penetration tests (CPT) typically relies on empirical correlations between the soil cyclic resistance ratio, CRR and cone tip resistance qt. For sands that containˆnes (particles passing #200 sieve), an adjustment based onˆnes content in the CRR-qt correlation is called for in currently available procedures. Earlier laboratory calibration tests in reconstituted silty sand specimens performed by the authors have indicated that partial drainage in CPT played an important role in the CRR-qt correlations. A series of laboratory tests on undisturbed samples andˆeld piezocone penetration tests (CPTU) were carried out in an alluvial soil deposit at two test sites in Central and Southern Taiwan. The results indicated that unlike the uniformly mixed specimens reconstituted in laboratory, the tested natural alluvial silt/sand soils can be heterogeneous with closely spaced clean sand layers embedded in a matrix of silty material. This stratiˆcation can render the CPTU a drained test even atˆnes contents well in excess of 50z. The ignorance of free draining eŠects on CPTU due to stratiˆcation in alluvial soils may lead to overestimation of CRR. Because of these factors, calibration of the CRR-qt correlations by performing tests in local soil andˆeld procedures to ascertain the drainage conditions in CPTU are important to the proper application of the CRR-qt correlations in the liquefaction potential assessment.