The first selection step in sugar cane breeding occurs on individual plants coming from true seeds. It is useful to assess the magnitude of uncontrolled factors affecting the rate and the efficiency, of the first selection. The effects of some aspects of competition on individual and family selection of seedlings have been analysed. The selection rate of 152 families representing, 9899 seedlings has been studied depending of the presence of dead (or very poor) neighbour seedlings. The average selection rate at ten months based mainly on the vigour of the plant was 21.5%. Depending on the presence of 0, 1 or 2 dead neighbours on the line the selection rate of the survivors was respectively 22.3%, 37%, and 41.1%. Fourteen families of 60 seedlings were observed for tillering. diameter of the stalk and Brix in a family selection trial. The seedlings were analysed depending on their location in the plot, e.g. near a border or in the middle of the plot. It was found that the location of the seedlings had an influence in the variance of the tillering but not in the other factors studied. The variance in tillering of the border plots was increased by 36% compared to the variance of the middle plot. Some improvement of the experimental designs is proposed. It is suggested that family selection be used with square plots of sufficient size and with controls known to have similar growth habit to seedlings such as vitroplantlets to minimize the competition effects. The evaluation of seedlings in ratoon as well as growing them in high density sowing boxes which maximize the competition should be avoided.