Relationships were determined for six chrysanthemum cultivars between the number of short days (SD) required for floral commitment and the average daily light integral. This was done to improve the timing of interrupted lighting, a technique involving the intercalation of a period of long days (LD) into the SD phase of growing. Thirteen greenhouse trials were conducted at various times of year, some with shading, at a minimum maintained temperature of 17-degrees-C. The basipetal commitment to flowering in SD was determined by transferring plants at intervals from SD to LD and subsequently comparing numbers of leaves below the flower bud on the main stems and lateral shoots of the transferred plants with those on other plants maintained continuously in SD. Numbers of SD required for complete commitment were then related to light receipt in the greenhouse. Whilst the speed of commitment varied considerably with cultivar, the form of the relationship for each was similar; asymptotic above 1.0-1.5 MJ m-2 day-1 photosynthetically active radiation (approximately 4.6-6.9 mol m-2 day-1), and exponential at lower average daily light integrals. The main effect of low light was to slow the commitment of the terminal meristem; subsequent commitment of axillary buds appeared to be independent of light integral. Confidence bands were calculated around the lines of best fit and relationships based on the upper limits of these have been used subsequently in conjunction with light monitoring for advising when to begin interruption. Analysis of historical light records for Little-hampton indicate the degree of variation to be expected from year to year for a given cultivar starting SD on a particular date during winter. Some practical aspects of using light-integral monitoring for the timing of interrupted lighting are discussed.