Accurate tumor staging depends on finding all tumor sites, and curative surgery requires the removal of all cancerous tissue from those sites. One technique for locating tumors is to inject patients before surgery with a radiotracer that is preferentially taken up by cancerous tissue. Then an intra-operative gamma-sensitive probe is used to locate the tumors. Small (< 1 cm diameter) tumors can be found with probes. These tumors are often undetectable by external imaging and the standard surgical inspection with sight and touch. Simple calculations and model tumor tests show that small tumors should be detected by single-element probes, but often such probes fail to detect them in practice. This discrepancy is often caused by the use of a uniform background to predict probe performance. Real backgrounds are nonuniform and the variations in the background decrease probe performance dramatically. We devised a method to predict probe performance in a realistic background which includes variations in normal organ uptakes. We predict the relative performance of probes, both existing devices and those still in the design stage. Since probe configurations can be compared without actually building devices, optimal detector and collimator configurations can be determined. The procedure uses a Monte-Carlo-calculated point response function, a numerical torso phantom (provided by George Zubal of Yale University) obtained from computed tomography slices of a normal male, and estimated biodistribution data derived from actual biodistributions of a monoclonal antibody. The Hotelling Trace Value, a measure of the tumor-detection performance, is computed from the probe responses at 5000 simulated sites. We present the relative detection performances of single-element, dual-element, imaging, and coincidence probes and determine the optimal configuration for collimation of an imaging probes.