Due to the uneven and often inadequate quality of cross-table lateral hip radiographs, many radiographic studies of femoral head penetration into polyethylene in total hip arthroplasty are limited to the two-dimensional measurement of femoral head penetration using the A/P film only. We postulated that the use of two oblique frontal projections at 90 degrees to each other would improve the three-dimensional evaluation. Using an established hip phantom, the idealized accuracy and precision of the three-dimensional Martell method was evaluated, contrasting the standard A/P and cross-table lateral projections versus a pair of oblique projections by four independent readers. Accuracy and precision resulting from the use of two oblique projections (average accuracy +/-63 microm, precision +/-26 microm) were similar to that obtained using the conventional A/P and cross-table lateral views (accuracy +/-54 microm, precision +/-22 microm), though the results of the two oblique views were slightly more variable. These observations suggest that by using two oblique A/P projections, the major disadvantage of using the cross-table lateral films, namely the variable quality of the images, is avoided. Perhaps, therefore, the utility and availability of three-dimensional data in comparable clinical studies may be improved.