The paper deals with a relatively new concept in wave-energy conversion systems: the Anaconda. This device consists basically of a distensible tube, made out of rubber material, which interacts with the incident waves in order to convey the absorbed energy to some form of power take-off (PTO). In the present case, the bulges inside the rubber tube drive a slug of water towards a vertical shaft referred to as a chimney. The oscillations of the water column inside the chimney induce a pressure oscillation in the pneumatic chamber above that activates an air-turbine. The Anaconda device has been studied extensively by experimental and theoretical means. However, so far, the pneumatic PTO for this device has only been examined within restrictive linear conditions. This study is about a series of 1:100 scale model tests of a freely floating Anaconda in a wave flume, under deep and intermediate water regular waves, for a device equipped with a non-linear PTO. In the model, the pneumatic chamber communicates with the atmosphere through an orifice. Three calibrated orifices of different diameters have been tested. The pressure across the orifice is determined from measurements taken of the amplitude of water column oscillations in the system, by applying Bernoulli’s law. These properties are subsequently used for calculating the power captured from the incident waves. Estimates of power output and energy capture width are presented, as functions of wave frequency, while the device interacts with linear and second-order waves. Our estimates are finally extrapolated to predict prototype performance.