The short acquisition time of single-shot Echo Planar Imaging (ssEPI) and its ability to freeze intra-slice motion makes it ideal for both functional (fMRI) and diffusion (dMRI) studies in the fetus. Its efficiency depends both on in-plane acceleration, the speed of k-space transversals and the time for any contrast weighting which needs to be performed for every slice. Both the long diffusion preparation and the longer echo times used in fMRI to maximize BOLD contrast contribute to inefficiency in ss-EPI acquisitions. Beside Multi-band (MB) imaging, which reduces the total acquisition time by acquiring multiple simultaneous slices and using the coil sensitivities to unfold, multiplex (MP) imaging has also been proposed [1,2]. Here, multiple slice locations are excited by sequential excitation pulses separated by a frequency offset, and read-out interleaved within an EPI echo read-out train (Fig. 1). The time required for the additional encoding and excitation pulses leads to a penalty in echo time (TE) and EPI readout duration. However, fetal imaging, has imposed limitations on gradient switching due to acoustic noise considerations and PNS limits which increase echo spacing and thus may be an ideal candidate for multiplex imaging. This abstract demonstrates a combined multiband-multiplex strategy with MB refocusing pulse to refocus all excited locations, from both MB and MP slices. This removes the need to have a widened slice refocus pulse as originally proposed [1] allowing MP and MB slices to be interleaved with constant slice separation leading to a uniform final slice stack.To allow safe operation for fetal imaging, the acoustic noise generated by the sequence was reduced using sinusoidal read-out at a fixed frequency, a constant phase encoding gradient and a strategy to combine the butterfly refocusing crushers with the diffusion gradients (Fig. 1) is used. Furthermore, the slice refocusing gradients between both excitation pulses are combined to avoid sharp gradient rises. In total, MBxMP individual slices are excited and read-out within each EPI read-out. While the excitation pulses may be single-band or MB pulses, in this implementation the refocusing pulse is multiband accelerated with a factor of MB x MP. All excited slices are separated by a fixed gap of FOV (slice direction)/(MPxMB) (see Fig. 2a). To reduce pulse extension due to peak b1 demands, the optimal phase combination which produces amplitude (only) modulated pulses is used [3]. To limit TE difference between the two multiplex excitations, asymmetric pulses are used, the second being time reversed to minimise slice excitation time differences. For read encoding, the gradient area between excitation pulses (Gc) is used to control the separation of the 2MP k-space centres (separation shown in green in Fig. 2b), the pre-winder gradient (Gp) is adapted accordingly to cover the desired resolution (red in Fig. 2b). While the standard setting for two maximally separated fully acquired k-spaces means Gc = 0.5 K and Gp = 0.25 K`, as illustrated in the upper half of Fig. 2B, higher resolution from partial echo in read-direction is possible (lower half). The diffusion-weighted ssEPI sequence was tested on a clinical 3T Achieva scanner (Philips, The Netherlands) on two fetuses (gestational ages 30+1, 34+2) after informed consent was obtained. Parameters included: FOV = 315mm x 315mm x 77mm; resolution of 3.2mm isotropic; b-values of b=0, b=100, b=400; TE=115ms, TR=6000ms, no SENSE, halfscan = 0.838. Singleband (MB1-MP1), Multiband2 (MB2-MP1), Singleband-Multiplex2 (MB1-MP2), Multiband4 (MB4-MP1), and Multiband2-Multiplex2 (MB2-MP2) versions were acquired and reconstructed using in-house reconstruction algorithms.The results for the singleband, the 2-fold accelerated (m2-s1 & m1-s2) and the 4 fold accelerated (m2-s2 & m4-s1) acquisitions are shown in Figure 3. Despite the challenges of fetal imaging - fetal and maternal motion, as well as maternal fat - it does demonstrate the ability to obtain combined multiband-multiplex accelerated images of the fetal brain. Fig. 4 shows a diffusion weighted study (b0,b100,b400) acquired with the implemented multiplex imaging capacity, reducing the imaging time by roughly half. As there was no penalty on the EPI read-out train due to inherent limitations in fetal imaging, the only efficiency loss is due to the duration of the second excitation pulse and the prolonged MB4 refocusing pulse, otherwise multiplex2 imaging allowed almost double the speed of the acquisition. Multiplex methods can combine synergistically with other slice acceleration techniques such as partial Fourier and SENSE. Further work will include partial echo in read direction as described above as well as improvements in the reconstruction.