Quantification fatty liver is useful for early detection of liver disease. Now ultrasonic detection is used for screening fatty liver. But it lacks of quantity and objectivity. Photoacoustic (PA) imaging, a novel modality, is expected for applying to estimate fat rate of liver. To realize estimation of fat rate, we developed handheld PA system using multi wavelengths. In this study, we verified the usefulness of the system for estimate fat rate in ex vivo experiments. As a biological sample, the mixtures of chicken liver, olive oil and lard (the mixing rate of the lipid in the liver is from 0 to 0.3with an interval 0.1) was used. We acquired PA spectra of the sample by developed handheld system. Nanosecond pulses of laser light (800-1300 nm wavelength, 30 Hz repetition rate) were guided to the sample surface by an optical fiber bundle close to the linear ultrasound probe. By analyzing PA spectra, we can find liver have peak PA intensity around 900 nm and lipid have it around 1210 nm where the light absorbance is high. To estimate fat rate using photoacoustic methods, the photoacoustic signal intensity ratio between two wavelength regions was calculated as described above. Signal intensity ratios agreed well with the composition ratio between liver and lipid. From these analyses, the advantage of PA spectra for estimation of fat rate, and the feasibility of our system for early detection of fatty liver were demonstrated. Devising a technique for quantifying fatty liver and in-vivo experiments will be conducted in further studies.