Neuro-electrophysiology allows scientists to investigate the underlying electrical properties that constitute brain neural network assembly. Developing tools to study these properties is a rapidly-evolving research field, and recent advancements in micro electrode arrays (MEAs) is opening a new frontier in long-term data acquisition. MEA microfabrication techniques have advanced over the years and led to different types of electrodes. The objective of this study was to optimize MEA design featuring multiple wells per electrodes (MW-MEA), to improve the recording efficiency of MEAs used for in vitro electrophysiological recordings. Methods: Two multi-well electrode designs (5 wells with diameter of 20 μm vs. 6 wells with diameter of 15 μm) were evaluated. Peak to peak signal amplitude of the recorded signals and the noise levels were studied and the signal to noise ratios (SNR) were determined. Results: The signal amplitudes recorded by electrodes with 6 wells (1060.3 μm2 ) was higher than those recorded by 5-wells electrodes (1570.8 μm2 ), while the noise level remained identical in both designs (31.3 μv ± 10.2). As such, the SNR recorded by the 6-wells electrodes showed a 1.8-time increase, compared to the electrodes with 5 wells, although the diameter of the wells in the former design was smaller. The results of this study demonstrated an inverse relation between SNR and open surface area of electrodes. Significance: The identified relation between electrode well characteristics and MW-MEA performance and design optimization can improve signals’ resolution during long-term spontaneous extracellular recordings and thus the quality of brain cell activity recordings.