Footwear sole wear by natural use or artificial abrasion either increases or decreases slip resistance on floors with and without lubricant. The objectives of this paper were to study the effect of footwear sole abrasive wear on slip resistance on ice with respect to temperature, and to compare the slip resistance of abraded soles on melting and hard ice with that on lubricated steel plate. The kinetic coefficient of friction (COF) of nine pairs of footwear were measured with the stationary step simulator developed at the Finnish Institute of Occupational Health, before and after the new footwear soles were artificially abraded. Two-way factorial ANOVA showed that the abrasion of nine pairs of footwear had no significant effect on COF on melting ice (Mean COF with abrasion=0.056, std=0.0158, COF without abrasion=0.055, std=0.0205, P=0.805). On hard ice, however, the COF of abraded soles measured (mean COF=0.244) was significantly higher than without abrasion (mean COF=0.180, p<0.001), and than abraded soles on lubricated steel (mean=0.137, p<0.001). There is statistical significance between the three types of surfaces (p<0.001). On hard ice, regardless of abrasion, curling footwear with crepe rubber soling showed significantly higher COF (mean=0.343 after abrasion, 0.261 before abrasion) than other types (p<0.001). The results indicate that artificially abraded footwear is more slip resistant than new one for use on hard ice. The abrasion requirement could be specified if developing a new standard to measure COF on ice in the future. Of the footwear measured, the curling footwear with crepe rubber soling performed best in terms of slip resistance property. Therefore, Crepe rubber soling is highly recommended for use on hard ice. Melting ice is much more slippery, in which sole abrasion does not improve slip resistance. Thus, additional measures should be taken to reduce slip and fall risk.