An accumulated soil temperature model is described which calculates the first oviposition date for overwintering populations of the cecidomyiid Resseliella theobaldi. Data for the model were collected from observations of the oviposition behaviour of R. theobaldi in a raspberry (cv. Glen Cova) plantation in the UK over an 8-year period, from 1977 to 79 and 1981 to 85. The model used values for the mean 10-cm soil temperature derived from the daily maximum and minimum air temperature and the 10-cm soil temperature at 09.00 h. The model was developed by minimising variation between the observed dates for the start of oviposition over the 8-year period and the accumulated air or soil temperature using a series of base temperatures. The best fit was obtained with soil temperatures accumulated above a base temperature of 4C. Oviposition was predicted to start when the daily accumulated soil temperature reached 339 day degrees C above a base temperature of 4C. Field observations and model predictions suggested that the emergence and oviposition dates of R. theobaldi were influenced by the direction and slope of the raspberry plantation. To compensate for this, an empirical correction was incorporated into the model to advance or retard the predicted oviposition date by adjusting the estimated maximum soil temperature to allow for the slope. The geographical variation between the dates for the first oviposition was simulated by running the model for 9 meteorological sites in the UK in 1985 and 1986.