This study adds a new dimension to lab-on-a-chip systems by employing three-dimensional (3D) integration technology for improved performance, higher functionality, and on-chip computational power. Despite the extensive amount of current research on 3D memory modules, microprocessors, and FPGAs, manufacturing and packaging challenges, as well as cost and reliability concerns have prevented the development of a 3D integrated lab-on-a-chip system. The aim of the present study was to demonstrate the feasibility of individual steps that could pave the way for the realization of smart, autonomous, low-cost, and compact biosensors for applications such as genome research, point-of-care diagnostics, drug discovery, and cell manipulation and sensing. Undoubtedly, such integration requires broad and multi-disciplinary research that combines microfabrication, biosensing, circuit design, and testing. This thesis addresses all of these aspects from manufacturing, reliability, and cost perspectives. Three major work packages are presented: (i) disposable biochip with flexible microcontacts; (ii) through-silicon-via (TSV) fabrication and chip-to-chip (C2C) integration; and (iii) microelectrode and circuit development for biosensing. Integrating the biosensing function and microfluidics on top of CMOS electronics has enabled a new generation of lab-on-a-chip systems that have addressing, sensing, and data elaboration functions on the same device. However, a key parameter for the widespread use of such system in the biosensing field is the disposability of the assay-substrate. In this context, the present study proposes a disposable biosensing layer that can be aligned and temporarily attached to the electronics through flexible interconnections and can be replaced after each measurement to eliminate the cleaning steps and cross-contamination of samples. This idea merges the advantages of passive- and active-electronic biosensors in one system and promises three key benefits: (i) high-density microelectrode array thanks to vertical interconnections; (ii) high-performance operation thanks to circuits in close proximity; and (iii) low-cost, disposable, and configurable biochips by fully decoupling the fabrication of the sensor and the electronics. The replaceable biosensing layer developed in this study is fabricated by employing four photolithography steps and by avoiding laborious and costly processes. Flexible microcontacts are realized by metal patterning and parylene deposition inside the trenches, etched on a silicon wafer by DRIE and KOH. Novel techniques are developed to pattern the frontside micro-electrode arrays and seal the wafer surface in order to prevent leakage during the measurements in liquid environment. Backside silicon DRIE is employed to form the openings for CMOS chip placement, where the alignment accuracy by manual chip placement was measured as 5 μm on average. During the preliminary tests, mechanical and electrical contact measurements were conducted by placing a dummy chip into the backside opening. Estimations of the fabrication cost, design and process issues, and alternative approaches were also provided. Regarding the 3D integration part, this thesis presents three different post-CMOS processing techniques: (i) TSV fabrication based on stencil lithography and bottom-up Cu electroplating; (ii) C2C integration based on wafer reconstitution from diced chips, standard photolithography, and template alignment; and (iii) post-processing based on dry-film lithography, C2C alignment by surface-tension-driven self-alignment, and parylene bonding. Daisy-chain resistance measurements demonstrated 0.5 Ω of TSV resistance and 99% yield for 1280 TSVs. Moreover, die-level post-CMOS processing was successfully demonstrated with real microprocessor chips for 40 μm diameter via etching, redistribution layer patterning, and C2C bonding. The study has verified that the proposed techniques offer a simple and low-cost solution for bulk and surface micromachining on diced chips, particularly when compatibility issues and the cost of the CMOS make it impossible to process the full-wafer. This study also shows that 3D integration, which is considered to be a very challenging process, can be accomplished in a MEMS cleanroom without using expensive specialized tools. Tethered bilayer lipid membrane (tBLM) and electrochemical-based biosensors were chosen as test vehicles for the biosensor part. Passive microelectrodes and control/readout electronics were designed and implemented. Membrane resistance change in response to ion channel formation in tBLM biosensor was measured in real time using a switched-capacitor integrator-based interface circuit. In the latter biosensing technique, redox species on the microelectrode surface were detected through cyclic voltammetry and square-wave voltammetry methods. The reliability of on-chip microelectrodes employed in aqueous solutions and subjected to chemical surface modification was inspected by conducting a comparative study on the quality of different passivation layers, such as sputtered SiO2, LPCVD low-temperature-oxide, Parylene-C, SU-8, and dry-film resist. This thesis presents detailed fabrication processes for each technique, their advantages and disadvantages, as well as statistical analyses of the impedance measurement results of 300 electrodes. Having verified the stability of the microelectrodes through impedance measurements, the study achieved reproducible results on electrochemical label detection. Finally, the thesis discusses the future work towards full system integration.