When a scanning tunnelling microscope (STM) tip is driven into a metallic sample surface a nanometer sized wire (nanowire) is formed during the subsequent retraction. The electrical conduction measured during this retraction process shows signs of quantized conductance in units of 2e(2)/h. Due to the inherent non-reproducibility of the measured conductance curves a standard technique is to build histograms from a large number of curves. Such histograms, built with conductance experiments on gold nanowires at room temperature, show 3-4 peaks at integer values of 2e(2)/h, while in a low temperature mechanically controlled break junction study only the first peak is reported. In this work, histograms made up of thousands of consecutive curves at 4K are presented, showing up to 5 conductance peaks. An explanation for this discrepancy could be a higher nanowire temperature resulting from the higher retraction speed used in our measurements. However, a simple estimation, where we used macroscopic heat transport theory, resulted in a very low temperature increase, less than 1 mu k, ruling out this possibility. Thus, no significant difference with previous room temperature studies were observed, pointing to a conductance quantization that is the same at room and low temperature.