Three multimicrophone probe arrangements used to measure acoustic intensity are the four-microphone regular tetrahedral, the four-microphone orthogonal, and the six-microphone designs. Finite-sum and finite-difference processing methods can be used with such probes to estimate pressure and particle velocity, respectively. A numerical analysis is performed to investigate the bias inherent in each combination of probe design and processing method. Probes consisting of matched point sensor microphones both embedded and not embedded on the surface of a rigid sphere are considered. Results are given for plane wave fields in terms of root-mean-square average bias and maximum bias as a function of angle of incidence. An experimental verification of the analysis model is described. Of the combinations considered and under the stated conditions, the orthogonal probe using the origin microphone for the pressure estimate is shown to have the lowest amount of intensity magnitude bias. Lowest intensity direction bias comes from the six-microphone probe using an average of the 15 intensity components calculated using all microphone pairs. Also discussed are how multimicrophone probes can advantageously use correction factors calculated from a numerical analysis and how the results of such an analysis depend on the chosen definition of the dimensionless frequency.