Since its introduction the PinPoint (PTW-Freiburg) micro-ionization chamber has been proposed for relative dosimetry (output factors, depth dose curves, and beam profiles) as well as for determination of absolute dose of small high-energy photon beams. This paper investigates the dosimetric performance of a new design (type 31014) of the PinPoint ion chamber with a central aluminum electrode. The study included characterization of inherent and radiation-induced leakage, ion collection efficiency and polarity effect, relative response of the chamber, measurement of beam profiles, and depth dose curves. The 6 and 15 MV photon beams of a Varian 2100 C/D were considered. At the nominal operating voltage of 400 V the PinPoint type 31014 chamber was found to present a strong field size dependence of the polarity correction factor and an excess of the collected charge, which can lead to an underestimation of the collection efficiency if determined with the conventional "two-voltage" method. In comparison to the original PinPoint design (type 31006) the authors found for type 31014 chamber no overresponse to large-area fields if polarity correction is applied. If no correction is taken into consideration, the authors found the chamber's output to be inaccurate for large-area fields (0.5% accuracy limited up to the 12 x 12 and 20 x 20 cm2 field for the 6 and 15 MV beams, respectively), which is a direct consequence of the stem and polarity effects due to the chamber's very small sensitive volume (0.015 cc) and cable irradiation. Beam profiles and depth dose curves measured with type 31014 PinPoint chamber for small and medium size fields were compared to data measured with a 0.125 cc ion chamber and with high-resolution Kodak EDR2 films. Analysis of the penumbra (80%-20% distance) showed that the spatial resolution of type 31014 PinPoint ion chamber approaches (penumbra broadening < or = 0.6 mm) EDR2 film results.