periodically modulated in and out an absorption line of the studied molecule. During the relaxation process, energy is mainly transferred by collision to the surrounding molecules. A local change of temperature is then induced along the laser beam, and a standing wave is created inside the cell.Photoacoustic cells are often used in resonant mode in order to take advantage of the enhancement of the acoustical signal. The resonator shape directly affects the strength and the distribution of these pressure variations. To take full benefits of the sound wave amplification, the quality factor, the resonant frequency and the peak amplitude at resonance are the three main parameters to be determined. These three parameters are usually experimentally determined from the cell response:U is the measured voltage of the microphone (V), R M its sensitivity (V/Pa), W the optical power of the laser (W), c the gas concentration and K the absorption coefficient (cm −1 ). The maximum value of R c (ω) is often referred as the cell constant. This constant is used to determine the detection limit of the photoacoustic spectrometer. In order to enhance this detection limit, one must be able to predict the cell response.Positions of resonant frequencies and Q-factors can be calculated for simple resonant volume where analytic solution exists [2]. For more complex cells, electric analogy can also be used to investigate the cell response [3]. Finite element method (FEM) presents the ease of use of a computational calculation and has already demonstrated its capabilities for the simulation of the photoacoustic cell characteristics. Baumann et al. [4] predicted the resonant frequencies and Q-factors of a T-shaped cell with a goodAbstract We report the complete designing process and realization of a photoacoustic spectrometer. In a first step, the cell design is optimized in order to achieve maximum cell constant and working frequency using a finite element method. Technological and integration constraints are used to define dimensional constraints on the cell. In a second step, a dedicated optical bench is presented along with the photoacoustic cell. The resonator response is then measured using a quantum cascade laser for methane detection and condenser microphones as detectors. The system detection limit is also discussed as it depends not only on the cell response but is also a combination of parameters linked together: environmental noise, microphones characteristics and cell conception. The gas flow required in a dynamic operation of the sensor degrades the detection limit regardless of the microphones quality. Choices on cell conception to minimize gas flow noise are discussed.