The sliding velocity on ice, as opposed to coefficient of friction measurements, is proposed to show the effect of experimental conditions on the sliding velocity. A sled was positioned with one runner in a groove and the other on flat ice and then released to slide 25 m down an ice track. The sliding velocity was calculated from 20 repeat experiments using runners with a different roughness (polished vs abraded) under different loads (30 kg, 40 kg and 50 kg) and at a range of air temperatures (−15 °C, −10 °C, −2 °C, 0 °C and 2 °C). The error in sliding velocity was sufficiently small (0.15%) to clearly show the effect of different parameters, and allowed trends to be identified. Sliding velocity was mostly influenced by a change in temperature, then by a change in load, and then the runner roughness. The highest speed was recorded at 0 °C, when the roughness and intermediate loads played a larger effect on the sliding velocity. The results are discussed in reference to solid friction, movement over a water film, and the mechanical properties of the ice. Measuring the sliding velocity, with a controlled release mechanism on the sled obviates the operated related error, and also provides a good insight on the influence of different conditions, instead of measuring the ice friction and making predictions of the acceleration and the resulting sliding velocity.