The fungus, Sclerotium cepivorum Berk. is the causal agent of onion white rot disease and reproduces solely by means of sclerotia that persist in the soil between susceptible crops. Sclerotia are, therefore, a vulnerable part of the lifecycle and disease control measures are being targeted towards them. These methods depend on a thorough knowledge of sclerotial biology. Three key aspects of sclerotial biology were investigated in the course of this research programme; the factors that affect survival of sclerotia in soil, the phenomenon of sclerotial dormancy and the relationship between inoculum levels in the soil and subsequent disease incidence. The influence of soil type, location and sclerotial condition on survival of sclerotia was examined under field conditions. There was no significant difference in sclerotial survival in the two soil types tested (Patumahoe clay loam and Wakanui silt loam). Experiments were conducted at two locations (Auckland and Lincoln), which differed substantially in both average soil temperature and rainfall; Auckland having a warmer wetter climate. Location significantly affected sclerotial survival over time, with fewer sclerotia surviving at Auckland. The condition of the sclerotia, related to structural damage caused by desiccation and re-wetting in the field, was an important influence on survival. Most sclerotia (40 - 60%) decayed after just two months in soil, after which numbers remained relatively stable for up to 2 years. During the growing season, sclerotia became progressively more contaminated by other fungi and exhibited signs of desiccation. The sudden drop in viability over the first few months in soil was attributed to damage caused by adverse environmental conditions and subsequent attack by microbes. A study was made of the dormancy requirements of naturally-produced sclerotia in New Zealand soils. Dormancy was tested in sclerotia from 5 isolates buried in one of two different soils (Patumahoe clay loam and Wakanui silt loam). Two established methods were modified to test for dormancy in New Zealand conditions; one based in the laboratory and one in the field. In the laboratory-based experiment, < 40% of sclerotia germinated after 15 weeks in soil. Soil type did not significantly affect the length of dormancy in this study. The low rates of germination observed resulted in no significant differences among the isolates in the length of dormancy. This method was prone to problems caused by contamination by other fungi and is not recommended for use in New Zealand conditions. In the field trial, 47 - 72% of sclerotia had germinated or decayed after 4 months in soil, depending on the isolate. Sclerotia required 6 months in soil before high rates of germination occurred (89 - 98%). In the field trial, isolates differed in the amount of time required in soil before dormancy began to break (3 - 5 months). In both experiments described above, dormancy lasted several months longer than reported in the literature. Sclerotia collected from onions in a commercial field also exhibited sclerotial dormancy. This has implications for the use of germination stimulants (e.g. diallyl disuphide, DADS) in disease control in New Zealand. With dormancy lasting up to 6 months in some circumstances, a spring treatment of a germination stimulant would seem more appropriate than an autumn treatment following an outbreak of white rot. This is in line with overseas recommendations and should be practical as DADS can be applied while a cover crop is in place without any adverse effects. It may be necessary to delay planting until 8 weeks after the last DADS treatment, to ensure that seedlings are not exposed to a burst of germinating sclerotia. Eleven field trials were conducted over three years. The aim of these was to describe the relationship between inoculum density (ID) in the soil at the time of planting and incidence of white rot at harvest. Soil samples were taken from each trial site and the numbers of sclerotia present in them determined. A novel method was developed to separate sclerotia from soil. This technique used the magnetic properties of haematite, an iron ore common in the soils of the Pukekohe region, to divide soil particles from sclerotia. Disease incidence, soil temperature and moisture were then recorded through the growing season. Curve-fitting was performed to determine the pattern of disease progress. In the first growing season (1998 / 99) data from 57% of the quadrats were sucessfully fitted to the gompertz model, one of the sigmoid curves. In the second year (1999 / 2000), low ID (and disease incidence) meant that only 26% of quadrats could be fitted to the gompertz model. In the third year, the pattern of disease was unlike that of the previous two seasons, with lower soil temperatures probably the cause of a later onset of white rot and continuous increase in disease incidence until harvest. In previous years, numbers of new infections had decreased in November - December. In 2000 / 01, the rate of new infections continued to increase into December. Data collected during the third year was fitted to exponential, log and linear models, with all data fitting one of these models. There was a positive correlation in year one between inoculum density and time to maximum rate of disease increase (TMR), with high ID resulting in an earlier disease onset by approximately 4 weeks. A positive, but not statistically significant, trend was also observed between ID and both maximum rate of disease increase (MR) and the point of inflection in the curve (LI). No correlations between ID and these parameters were identified in 1999, due probably to the low levels of inoculum and disease observed. TMR, LI and MR were not calculated for the 2000 data, as these could not be fitted to the gompertz model.