1. A seed vigour test provides a more sensitive measurement of seed quality than the standard germination test. The cold test is one of the most widely used vigour tests for maize but procedures had not been standardized. In this study different cold test methods were compared for their usefulness as a seed vigour test. 2. Field emergence of 12 hybrid seed lots and 10 parent line seed lots were determined in four different field trials, representing different environmental conditions. Conditions were cold and wet, cool and wet, and favourable. Percentage emergence and emergence rates were determined for the different trials. 3. Considerable differences were found between percentage emergence in the different trials and the standard germination test results. The lowest emergence counts as well as emergence rates were obtained with the oldest seed lots. The differences found in emergence counts between the different seed lots appear to be the result of ageing and dete¬rioration and thus differences in vigour. The ranking order of the different seed lots according to percentage emergence were similar for the different trials. 4. Six different cold test methods and the standard germination test were conducted in the laboratory and the results of the different tests were correlated with field emergen¬ce. The deep-box and rolled towel methods were used and the germination substrates used were sand, a sand-soil mixture, vermiculite, a vermiculite-soil mixture and paper towels with and without soil. 5. Considerable differences were found in percentage emergence between the different laboratory tests. The lowest vigour percentages were obtained with the older seed lots, thus a decline in percentage emergence with an increase in age of the seed lot. The same tendency was found with the field emergence trials. The ranking order of the seed lots did not show large differences between the different tests. 6. Lower vigour percentages were obtained when soil was included in the germination medium and this effect was more pronounced with the older seed lots. The sand + soil cold test was found to be the strictest of the cold tests with the lowest percentage emergence. 7. Correlations were determined between the results of the different cold tests and emergence in the different field trials. All correlation coefficients of the hybrid seed lots were found to be significant at the 1 % level. However, the correlation coefficients obtained between the standard germination test and field emergence were generally of a lower order than those obtained with the cold tests. No significant correlations were 61 found between emergence and the germination percentages of the parent line seed lots. 8. Although the correlations between the different laboratory test results and field emergence do not provide ground for a clear recommendation of adoption of one cold test procedure over another, inclusion of soil in the germination medium resulted in a small increase in correlation coefficients. Correlation coefficients between emergence and the sand + soil cold test were the highest in most of the trials . 9. Although any of the cold tests which include soil in the medium appear to be suitable for adoption by South African seed laboratories, the vermiculite + soil test has advantages over the sand + soil and paper + soil tests. / Dissertation (MSc (Botany))--University of Pretoria, 2006. / Plant Science / unrestricted