A long-term study of the population cycles of Helicoverpa armigera (Hubner), Tortrix capensana (Walker) and Cryptophlebia leucotreta (Meyrick) on citrus at Zebediela was undertaken to establish a phenological model for the planning and implementation of control strategies. Helicoverpa armigera and T. capensana were monitored with light-traps and orchard scouting. In H. armigera the relationships between adult activity, oviposition on citrus blossoms, the presence of larvae on citrus trees and the correlation between moth activity and climatic factors were determined. A correlation between moth activity in autumn and pest status in the following spring was demonstrated. Although the oviposition levels on Navel blossoms were higher than on Valencia, both cultivars were subject to an oviposition peak in the first week of September. The larval population peaked on both cultivars in the third week of September during full bloom. Larvae appeared earlier on Navel and but persistedlonger on Valencia. In a simple regression analysis, light-trap catches correlated with minimum and maximum temperatures, evaporation, and wind and saturation deficit. In a multiple regression analysis a positive correlation was demonstrated between minimum temperature and wind, and moth catches. The packhouse culling figure owing to damage by H. armigera damage for 1986-1993 was 0.43 % (S.D. = 0.27) for Navel and 0.29 % (S.D. = 0.25) for Valencia fruit. Tortrix capensana activity in winter and the following spring were correlated through light-trap catches. The adult population with the highest economic impact on citrus fruit quality occurred in the fourth week of September. Larvae produced by these moths appeared two weeks later. Subsequent moth generations between December and May were of no economic importance. In a simple regression analysis, light-trap catches showed the highest positive correlation with ambient temperature. In a multiple regression analysis evaporation andrainfall correlated positively with moth catches. The packhouse cull averaged 0.64 % (S.D. = 0.68 %) on Navels and 0.52 % (S.D. = 0.56 %) on Valencias. Lemons were not damaged. Male populations of C. leucotreta were monitored with virgin-female traps. The highest moth activity occurred in the second week of February. The relationship between moth catches and climatic factors were investigated, as well as the pattern and number of fruit lost as a result of C. leucotreta in Navel and Valencia orchards. The highest single incidence of Navel fruit loss occurred in December and Valencia fruit loss in July. In a simple regression analysis, weekly moth catches showed a correlation with minimum and maximum temperature, evaporation and rainfall. In a multiple regression analysis maximum temperature correlated positively and evaporation negatively with moth catches. A relationship between moth catches and fruit loss was established. According to a regression between moth catches and Navel fruit loss, a weekly average of 10 males/trap/22 ha of trees per week resulted in a loss of 0.53 cartons of fruit/ha/week four weeks later. The average percentage crop loss due to false codling moth was 1.6 % in Navels and 0.3 % in Valencias.