This article presents a high-performance microcavity electric field sensor (MEFS) based on silicon chip-based thin film lithium niobate (TFLN) photonic integrated circuits, utilizing the Pound-Drever-Hall (PDH) detection scheme. The MEFS achieves a detection sensitivity of 5.2 μV/(m√Hz), which is two orders of magnitude better than previous lithium niobate electro-optical electric field sensors and comparable to quantum sensing approaches. The sensor has a bandwidth up to three orders of magnitude broader than quantum sensing methods and can measure fast electric field amplitude and phase variations in real time. The ultra-sensitive MEFS represents a significant step towards building electric field sensing networks and broadens the application spectrum of integrated microcavities. Electric field sensing is essential in basic science with applications ranging from detecting cosmic fast radio bursts to understanding quantum physics and lightning origin. It is also crucial for ensuring the smooth operation of modern society, including monitoring power system reliability, testing electromagnetic compatibility in semiconductor foundries, and enabling radar imaging for vehicles. Various physical effects, including electro-optical, piezoelectric, electrostatic force, and quantum effects, have been used for electric field sensing. Quantum sensing using Rydberg atoms or trapped ions has reached detection limits down to 5.5 μV/(m√Hz) and 0.2 μV/(m√Hz), respectively. However, these methods require complex systems for quantum state preparation and have narrow instantaneous sensing bandwidths. Electro-optical (Pockels) approaches are relatively simple and can achieve high sensitivity and large bandwidth simultaneously. Optical interferometers made of electro-optical active materials have been widely used for electric field sensing, with lithium niobate (LN) being a material of keen interest due to its large Pockels coefficient, long-term reliability, and low loss. LN sensors are photonic integration compatible and on-chip waveguides have been used to form these interferometers for electric field sensing. However, long waveguides and electrodes are needed to improve sensitivity, which limits the sensitivity scaling due to velocity mismatch between electric and optical fields. The large form factor may also limit the sensor spatial resolution. Interferometer-based sensors must operate around the quadrature point, and ambient fluctuations impose practical issues on them. Instead of long waveguides, light circulation in micro optical cavities may also be used to enhance sensitivity with a compact form factor. High-Q microcavities have driven chip-based high-coherence lasers, classical and quantum microcombs, and have been widely used for ultrasensitive sensing of nanoparticles, acceleration, force, and rotation. They are also used as references to narrow laser linewidth down to sub-10 mHz, laying the foundation for next-generation timekeeping systems. The PDH method converts frequency detuning from the cavity resonance into a steep error signal, which can be used for sensing. The PDH scheme and enhancedThis article presents a high-performance microcavity electric field sensor (MEFS) based on silicon chip-based thin film lithium niobate (TFLN) photonic integrated circuits, utilizing the Pound-Drever-Hall (PDH) detection scheme. The MEFS achieves a detection sensitivity of 5.2 μV/(m√Hz), which is two orders of magnitude better than previous lithium niobate electro-optical electric field sensors and comparable to quantum sensing approaches. The sensor has a bandwidth up to three orders of magnitude broader than quantum sensing methods and can measure fast electric field amplitude and phase variations in real time. The ultra-sensitive MEFS represents a significant step towards building electric field sensing networks and broadens the application spectrum of integrated microcavities. Electric field sensing is essential in basic science with applications ranging from detecting cosmic fast radio bursts to understanding quantum physics and lightning origin. It is also crucial for ensuring the smooth operation of modern society, including monitoring power system reliability, testing electromagnetic compatibility in semiconductor foundries, and enabling radar imaging for vehicles. Various physical effects, including electro-optical, piezoelectric, electrostatic force, and quantum effects, have been used for electric field sensing. Quantum sensing using Rydberg atoms or trapped ions has reached detection limits down to 5.5 μV/(m√Hz) and 0.2 μV/(m√Hz), respectively. However, these methods require complex systems for quantum state preparation and have narrow instantaneous sensing bandwidths. Electro-optical (Pockels) approaches are relatively simple and can achieve high sensitivity and large bandwidth simultaneously. Optical interferometers made of electro-optical active materials have been widely used for electric field sensing, with lithium niobate (LN) being a material of keen interest due to its large Pockels coefficient, long-term reliability, and low loss. LN sensors are photonic integration compatible and on-chip waveguides have been used to form these interferometers for electric field sensing. However, long waveguides and electrodes are needed to improve sensitivity, which limits the sensitivity scaling due to velocity mismatch between electric and optical fields. The large form factor may also limit the sensor spatial resolution. Interferometer-based sensors must operate around the quadrature point, and ambient fluctuations impose practical issues on them. Instead of long waveguides, light circulation in micro optical cavities may also be used to enhance sensitivity with a compact form factor. High-Q microcavities have driven chip-based high-coherence lasers, classical and quantum microcombs, and have been widely used for ultrasensitive sensing of nanoparticles, acceleration, force, and rotation. They are also used as references to narrow laser linewidth down to sub-10 mHz, laying the foundation for next-generation timekeeping systems. The PDH method converts frequency detuning from the cavity resonance into a steep error signal, which can be used for sensing. The PDH scheme and enhanced