The neural signals are low-frequency (mHz–kHz) and low-amplitude signals (μV–mV). Therefore, the amplifiers for these signals must be low-noise circuits. Additionally, the front-end amplifiers must reject interference due to the common-mode signals as well as electrode effects. Amplification techniques based on clocking and continuous-time approaches are described. The clock-based techniques include switched-biasing, chopper, and auto-zeroing methods. The traditional continuous-time circuit is the AC-coupled-operational transconductance amplifier-based neural amplifier endowed with capacitive feedback. The unavoidable trade-off between input capacitance and area of the chip against the gain of the amplifier can be relaxed. This is achieved when a clamped T-capacitor network replaces the feedback capacitor.