Portable wireless devices, like phones and laptops, are an indispensable part of daily life worldwide. These devices are transceivers, meant to transmit and receive on many different frequency bands, often simultaneously. As transmitting emit electromagnetic waves, viewed as a potential health threat, wireless transmitting devices are regulated on how much exposure a person experiences during their use. Regulatory agencies impose limits on exposure measures such as specific absorption rate (SAR) and power density (PD) to ensure portable wireless devices are safe for public use. Recent developments in the new generation of portable wireless devices, such as the fifth generation of the cellular system (5G), have introduced devices with multiple transmitters operating in the millimeter-wave (mmWave) frequencies.This research aims to provide a practical solution for proper measurement procedures for multiple-antenna portable wireless devices and mmWave frequencies. We start by determining the compliance of multiple-antenna portable wireless devices exposure with the minimum number of measurements. Evaluation of regulatory requirements for multiple-antenna devices is time-consuming and costly in measurements since there is a large set of excitation signals for measurements. Therefore, the development of proper measurement procedures for when there are multiple transmitting antennas operating at the same carrier frequency is of great interest. To this end, we next model peak spatial-average SAR of multiple-antenna portable devices using a simple mixed quadratic function of the transmitted excitation signal. This model can be used in the design of any communication systems with exposure constraints, to reduce exposure and improve overall performance. Finally we propose a PD measurement procedure for evaluating exposure safety in mmWave frequencies. We show that greater accuracy for PD as a proxy for temperature can be achieved by adapting the averaging area according to a simple prescribed process that uses the half-power beamwidth of the excitation signal. Moreover, we show that testing must happen over a multitude of distances up to the far-field of the device.