For interference-limited multi-user communications, many papers have demonstrated, in theory, the effectiveness of Rate Splitting Multiple Access (RSMA) in suppressing the interference and achieving better outcomes (w.r.t spectral efficiency, fairness etc.) than the conventional Space Division Multiple Access (SDMA) used in present-day standards. However, an experimental demonstration of RSMA's benefits is missing in the literature. In this paper, we address this gap by realizing an RSMA prototype using software-defined radios. For the two-user multiple-input single-output (MISO) scenario, we measure the throughput performance of SDMA and RSMA in nine different scenarios that vary in terms of the channel pathloss and spatial correlation experienced by the users. Emulating perfect channel state information (CSI) at the transmitter through unquantized CSI feedback, we observe that RSMA achieves a higher sum throughput (upto 57%) and more fairness than SDMA when the user channels have high spatial correlation. Similarly, emulating imperfect CSI through quantized feedback, RSMA - along with the above trend - also experiences a smaller sum throughput loss, relative to the unquantized case (37%, on average, as opposed to 44% for SDMA). These outcomes are consistent with theoretical predictions, and demonstrate the feasibility and potential of RSMA for next generation wireless networks (e.g., 6G).