For body imaging at ultrahigh field strengths, the use of a combined array with transceive dipole antennas and receive-only loop coils has demonstrated superior imaging quality [1]. When placed on the same axis, dipole antennas and loop coils are inherently decoupled. In this study, we explore the boundaries of parallel imaging by fabricating a combined array consisting of eight fractionated dipole antennas and 64 loop coils.The intended setup to distribute 64 loop elements around the body is indicated in figure 1a. The loops are distributed over 16 elements, where each element contains four loop coils in a row (4RX). Each loop was 58 x 90 mm and had an unloaded vs loaded Q-factor ratio of 10. Eight of these elements are equipped with a fractionated dipole antenna (1TX). Both element types are shown in figure 1b. During transmit, each receive loop is detuned at three locations where the fourth detuning location is realized by the high preamp decoupling impedance. Within one element, neighboring loop coil elements are decoupled by overlap. The overall decoupling performance is evaluated for a test setup where two elements are placed on a pelvis shaped phantom with realistic loading conditions (figure 1b). The coupling matrix for all elements in this setup has been determined using a vector network analyser. The efficiency of the transmit antenna may suffer from the presence of four loop coils. To test it, B1+ measurements on the same pelvis shaped phantom have been performed for an original fractionated dipole antenna (no receive loops underneath) and one of the new dipole antennas with four receive loops. The SNR for the combined setup of 2x4 receive loops for each transmit antenna has been compared to the setup with 2 loops underneath each antenna, again using the same phantom and the setup as indicated in figure 1b. One volunteer (33y, BMI 27, informed consent) has been scanned with part of the developed setup in a Philips Achieva 7T system (Philips Healthcare, Cleveland, USA). The full array could not be connected because the system is currently still limited to 32 receive channels. Therefore, 2 active and 2 passive 4-loop elements have been positioned at the anterior side and at the posterior side of the pelvis. At the lateral sides of this setup, four additional transmit elements are positioned. All transmit antennas were driven with 8x1 kW amplifiers, fed by a custom developed vector modulator.The coupling matrix for the setup as indicated in figure 1b is presented in figure 2. Clearly, almost all relations have a coupling level of -16 dB, with the obvious exception of neighboring, non-overlapped elements (loop1-loop5, loop2-loop6 etc.). These show values up to -7 dB for the outer loop coils that extend over the edge of the phantom and are therefore see less loading. Note that all these coupling values are still decreased by preamp decoupling resulting in an additional reduction of -8 dB. B1+ measurements presented in figure 3 show that the presence of the loop coils does not reduce the B1+ efficiency of the transmit elements. Sagittal gradient echo images for the setup as indicated in figure 1b are compared to a setup with one fractionated dipole antenna with two loop coils underneath [1] in figure 4a and b. Clearly, the eight loop coils have a more evenly distributed signal intensity and reach further into the phantom. This is confirmed by low FA SNR measurements in the transverse plane, where for the 2-loop setup an off-set plane is chosen as indicated by the green line in figure 4b). The SNR values at … cm depth are … for the eight loop array and … for the 2 loop element. Preliminary imaging results have been obtained for a healthy volunteer. T2w images were acquired (TR/TE=2500/90 ms, 0.5x0.5x3 mm3, TSE-factor: 9) but not with the SNR that was expected.We have presented an array design that shows remarkable performance on the bench and in phantom studies. Including preamp decoupling, the inter-element coupling is lower than -15 dB for all coil correlations. Transmit performance of the antennas is not compromised by the presence of the receive loops. However, the imaging performance on a human subject does not fulfill the expectations. Finding the cause of this discrepancy will be subject of further study.